Skip to Content

Sensors at North Temperate Lakes

Printer-friendly versionPrinter-friendly version
Issue: 
Fall 2012

Corinna Gries (NTL), Aaron Stephenson (NTL), Ken Morrison (NTL)

In collaboration with the Global Lakes Ecological Observatory Network (GLEON), NTL is currently maintaining six lake buoys and one terrestrial weather station. Each buoy is equipped with a thermistor chain measuring water temperature at one meter intervals, and measures basic weather parameters above the water (air temperature, wind speed and direction, relative humidity, etc.). Other parameters measured on some buoys include photosynthetic active radiation, dissolved oxygen, conductivity, chlorophyll concentration, and phycocyanin concentration. One buoy is equipped with moving sensors that traverse the water column continuously. A single buoy can deliver up to 40 distinct data streams for the deepest lake; in total, ca. 140 data streams at one minute intervals are recorded. The sensors are controlled by Campbell CR1000 dataloggers which also function as temporary data repositories for up to four weeks of data. The network of dataloggers is configured and data is retrieved from these loggers using Loggernet, Campbell Scientific proprietary software. Communication between the point-of-presence computer running Loggernet and the field-deployed Campbell dataloggers uses a 900MHz radio network based on Freewave full-spectrum radios. The data are pulled every hour from the data loggers onto the point-of-presence servers. For one season, (May through November) one buoy can collect up to 60 MB worth of data.


A DataTurbine server is running to allow for easy sensor health monitoring. The data are not currently streamed into the database directly by DataTurbine; instead the data logger files are manually uploaded into temporary tables. Database triggers then apply range checks before data are archived, with appropriate QA/QC flagging, in their respective final tables. A web application that accesses these final tables allows users to query, subset, and download the specific data they want.


Several aspects of this approach, plus various custom applications (e.g. Ziggy data loader), have been in operation at NTL for many years. We are now in the process of updating the system with NSF and Moore Foundation funding to Tony Fountain at CalIT2. Additionally, the GCE Matlab toolbox will be incorporated into our system to work seamlessly with DataTurbine for quality control of streaming data. Also, another buoy has recently been installed which communicates data between two DataTurbine servers, one on-site and one at the point-of-presence server, eliminating the need for a datalogger. It can be expected from these projects that by the end of 2013 the installation and maintenance of DataTurbine will become more user friendly. Currently unresolved issues of concern are related to tracking the type and make of sensors, calibrations, and other maintenance activities.