PARENT SESSION
Organized Oral Session 38: Sensors and sensor networks in ecology
Organizer(s): SD Wullschleger, RB Jackson, and TE Dawson
Wednesday, August 10, 1:30 PM - 5:00 PM, Meeting Room 516 A, Level 5, Palais des congrès de Montréal
Wireless networks for modeling of canopy and soil processes in forests.
Clark, James*,1, Flikkema, Paul 2, Hungate, Bruce2, Koch, George2, Koch, George2, Sillett, Steve 3, 1 Duke University, Durham2 Northern Arizona University, Flagstaff3 Humbolt State University, Arcata
ABSTRACT- Understanding microclimate effects on trees, including gas exchange, growth, and allocation, requires spatio-temporal models that ingest environmental data collected at fines scales across landscapes. WiSARDNet (Wireless Sensing and Relay Device Network) is an ad hoc wireless sensor network optimized for dense in situ spatio-temporal monitoring of environmental and ecosystems processes. WiSARDNet is a complete distributed sensing system that includes (i) an array of standard multi-channel passive and active probe interfaces (e.g., multiple channels of temperature, light, and soil moisture) in each sensor, (ii) extensibility to other probes via multiple interfaces, and (iii) a PC-based human/network interface for network monitoring and control. These features, combined with an energy-efficient hardware/software architecture and network protocol stack, as well as careful energy management in a weatherproof package, allow high-performance data collection for large-scale scientific field studies. Here we demonstrate the design of our network, and we summarize methods for assimilating network data in spatio-temporal models of canopy architecture, water, and CO2 exchange. Models make use of a discretized representation of space and time and hierarchical Bayes models for inference on parameters and underlying state variables. The principal goals of this talk are to assess the current potential for inference and prediction and to provide recommendations for future development.
Key words: wireless networks, forest dynamics, spatio-temporal models