Evaluation of wireless sensor networks (WSNs) for remote wetland monitoring: design and initial results

Abstract

Here, we describe and evaluate two low-power wireless sensor networks (WSNs) designed to remotely monitor wetland hydrochemical dynamics over time scales ranging from minutes to decades. Each WSN (one student-built and one commercial) has multiple nodes to monitor water level, precipitation, evapotranspiration, temperature, and major solutes at user-defined time intervals. Both WSNs can be configured to report data in near real time via the internet. Based on deployments in two isolated wetlands, we report highly resolved water budgets, transient reversals of flow path, rates of transpiration from peatlands and the dynamics of chromophoric-dissolved organic matter and bulk ionic solutes (specific conductivity)-all on daily or subdaily time scales. Initial results indicate that direct precipitation and evapotranspiration dominate the hydrologic budget of both study wetlands, despite their relatively flat geomorphology and proximity to elevated uplands. Rates of transpiration from peatland sites were typically greater than evaporation from open waters but were more challenging to integrate spatially. Due to the high specific yield of peat, the hydrologic gradient between peatland and open water varied with precipitation events and intervening periods of dry out. The resultant flow path reversals implied that the flux of solutes across the riparian boundary varied over daily time scales. We conclude that WSNs can be deployed in remote wetland-dominated ecosystems at relatively low cost to assess the hydrochemical impacts of weather, climate, and other perturbations.

Estimates of evapotranspiration from contrasting Wisconsin peatlands based on diel water table oscillations

“…Wireless sensor networks were embedded in both bogs (Watras, et al, 2014a;www.wetlands.gleon.org). The peatland nodes Figure 1.…”

“…As explained below, we did not attempt to estimate ET at some peatland nodes (indicted by white triangles in Figure 1) because their specific yield curves were unreliable. A floating E pan was also installed in each pond to estimate E from open water (Watras, et al, 2014a;. In addition to the peatland nodes, a single node was established in each bog pond using a stilling well made from 6″ ID PVC pipe to monitor changes in surface water elevation.…”

“…In the heavily forested Northern Highland Lake District (NHLD) of Wisconsin and upper Michigan, peatlands comprise roughly 20% of the district's 6,400 km 2 surface area (Magnuson, Kratz, & Benson, 2006). There are no comparable measurements for ET from NHLD peatlands; but preliminary data showed an average ET flux of 4.8 ± 0.5 mm/d (SE x ) during 10 consecutive summer days during which E from an adjacent bog pond averaged 3.4 ± 0.1 mm/d (Watras, et al, 2014a). Estimates of direct E from NHLD lakes based on either energy balances, mass transfer models, or floating E pans are in remarkably good agreement, all indicating a flux of approximately 3 mm/d averaged over the ice-free seasons of multiple years (Lenters, Kratz, & Bowser, 2005;Watras, Morrison, & Rubsam, 2016).…”

Networked lake science: how the Global Lake Ecological Observatory Network (GLEON) works to understand, predict, and communicate lake ecosystem response to global change

“…Resolving the roles that lakes play in landscape carbon cycling will best be advanced with improved models of organic carbon (OC) cycling . Understanding OC dynamics in bog lakes, where much of the OC in northern hemispheres is stored, requires a much better understanding of how their hydrology differs from almost all other lake ecosystems (Watras et al 2013). The future of water quality in lakes and reservoirs will require a better understanding of both the climatic and the nutrient effects on cyanobacteria (Brookes and Carey 2011, Carey et al 2012b, Cottingham et al 2015, O'Reilly et al 2015, Schaeffer et al 2015 as well as how invasive species affect biodiversity and ecosystem function.…”

A critical survey on marsupial robotic teams for environmental monitoring of water bodies

“…Another limitation of sensor networks is their difficulty to assure continuous communication and data storage. In the absence of these functional requirements sensor networks have shown to be successful in water monitoring in riverine [7] and wetland [8] environments. The conventionally low complexity of these kind of sensor platforms limits their payload.…”