HR: 1340h
AN: H13A-0983    [Abstracts]
TI: A Data Management Framework for Real-Time Water Quality Monitoring
AU: * Mulyono, E
EM: emulyono@horizon.csueastbay.edu
AF: Calif. State Univ. East Bay, Dept. of Math & Computer Sci., Hayward, CA 94542,
AU: Yang, D
EM: david.yang@csueastbay.edu
AF: Calif. State Univ. East Bay, Dept. of Math & Computer Sci., Hayward, CA 94542,
AU: Craig, M
EM: mitchell.craig@csueastbay.edu
AF: Calif. State Univ. East Bay, Dept. of Earth & Env. Sci., Hayward, CA 94542,
AB: CSU East Bay operates two in-situ, near-real-time water quality monitoring stations in San Francisco Bay as a member of the Center for Integrative Coastal Ocean Observation, Research, and Education (CICORE) and the Central and Northern California Ocean Observing System (CeNCOOS). We have been operating stations at Dumbarton Pier and San Leandro Marina for the past two years. At each station, a sonde measures seven water quality parameters every six minutes. During the first year of operation, we retrieved data from the sondes every few weeks by visiting the sites and uploading data to a handheld logger. Last year we implemented a telemetry system utilizing a cellular CDMA modem to transfer data from the field to our data center on an hourly basis. Data from each station are initially stored in monthly files in native format. We import data from these files into a SQL database every hour. SQL is handled by Django, an open source web framework. Django provides a user- friendly web user interface (UI) to administer the data. We utilized parts of the Django UI for our database web- front, which allows users to access our database via the World Wide Web and perform basic queries. We also serve our data to other aggregating sites, including the central CICORE website and NOAA's National Data Buoy Center (NDBC). Since Django is written in Python, it allows us to integrate other Python modules into our software, such as the Matplot library for scientific graphics. We store our code in a Subversion repository, which keeps track of software revisions. Code is tested using Python's unittest and doctest modules within Django's testing facility, which warns us when our code modifications cause other parts of the software to break. During the past two years of data acquisition, we have incrementally updated our data model to accommodate changes in physical hardware, including equipment moves, instrument replacements, and sensor upgrades that affected data format.
UR: http://www.sci.csueastbay.edu/cicore
DE: 1848 Monitoring networks
DE: 1871 Surface water quality
DE: 1872 Time series analysis (3270, 4277, 4475)
DE: 1895 Instruments and techniques: monitoring
DE: 4262 Ocean observing systems
SC: Hydrology [H]
MN: 2007 Fall Meeting