Search

A 11 year data set of spatially distributed snow water equivalent (SWE) was used to inform a quantitative, near-optimal sensor placement methodology for real-time SWE estimation in the American River basin of California. Rank-based clustering was compared with geographically based clustering (subbasin delineation) to determine the existence of stationary covariance structures within the overall SWE data set. The historical SWE data, at 500 × 500 m resolution, were split into 8 years of training and 3 years of validation data. Within each cluster, a quantitative sensor-placement algorithm, based on maximizing the metric of mutual information, was implemented and compared with random placement. Gaussian process models were then built from validation data points selected by the algorithm to evaluate the efficacy of each placement approach. Rank-based clusters remained stable interannually, suggesting that rankings of pixel-by-pixel SWE exhibit stationary features that can be exploited by a sensor-placement algorithm, yielding a 200 mm average root-mean-square error (RMSE) for 20 randomly selected sensing locations. This outperformed geographic and basin-wide placement approaches, which generated 460 and 290 mm RMSE, respectively. Mutual information-based sampling provided the best placement strategy, improving RMSE between 0 and 100 mm compared with random placements. Increasing the number of rank-based clusters consistently lowered average RMSE from 400 mm for 1 cluster to 175 mm for 8 clusters, for 20 total sensors placed. To optimize sensor placement, we recommend a strategy that couples rank-based clustering with mutual information-based sampling design. © 2013. American Geophysical Union. All Rights Reserved.

[1] A wireless sensor network (WSN) was deployed as part of a water balance instrument cluster across a forested 1 km2 headwater catchment in the southern Sierra Nevada of California. The network, which integrates readings from over 300 sensors, provides spatially representative measurements of snow depth, solar radiation, relative humidity, soil moisture, and matric potential. The ability of this densely instrumented watershed to capture catchment-scale snow depth and soil moisture distributions is investigated through comparison with three comprehensive gridded surveys and 1 day of detailed lidar snow data. Statistical analysis shows that the network effectively characterized catchment-wide distributions of snow depth, while offering a cost-effective, reliable, and energy-efficient means for collecting distributed data in real time. A temporal analysis of snow depth variability reveals that canopy cover is the major explanatory variable of snow depth and that under-canopy measurements persistently show higher variability compared to those in open terrain. An analysis of soil moisture shows lower variability at deeper soil depth and a correlation between mean soil moisture and variability for shallow soils. A three-phase design procedure was used to optimize the WSN deployment. First, as off-the-shelf performance of current WSN platforms for large-scale, long-term deployments cannot be guaranteed, statistics from a prototype deployment were analyzed. Two indicators of network performance, the packet delivery ratio and received signal strength indicator, showed that for our site conditions, a conservative 50 m node-to-node spacing would ensure low-power, reliable, and robust network communications. Second, results from the prototype were used to refine hardware specifications and to guide the layout of the full 57-node wireless network. Of these nodes, 23 were used actively for sensing, while the remaining 34 nodes were used as signal repeaters to ensure proper spatial radio coverage and robust network operations. Further analysis of network statistics is conducted during the third, operational, phase to validate system performance. © 2012. American Geophysical Union.