Smart sensor web design using dynamic data assimilation and optimal control
To validate space-borne measurements of soil moisture, such as those that will be provided by the Soil Moisture Active Passive (SMAP) mission, in real-time, we need to determine how to best use an available set of ground sensors to estimate the true mean of soil moisture in any given area with given spatial variations. Soil Moisture Sensing Controller and oPtimal Estimator (SoilSCAPE) is a wireless in-situ sensor network technology, developed under the support of the NASA ESTO/AIST program, to provide multi-scale validation of soil moisture retrievals from the Soil Moisture Active and Passive (SMAP) mission. The SMAP sensor suite is expected to produce soil moisture retrievals at 3 km scale from the radar instrument, at 36 km from the radiometer, and at 10 km from the combination of the two sensors. To validate the retrieved soil moisture maps at any of these scales, it is necessary to perform in-situ observations at multiple scales (ten, hundreds, and thousands of meters), representative of the true spatial variability of soil moisture fields.
The most recent SoilSCAPE network, in the California central valley, has been designed, built, and deployed to accomplish this goal, and is expected to become a core validation site for SMAP. The network, which is currently still being deployed, will consist of 150 sensor nodes, each connected to of 3-4 soil moisture sensors at various depths, deployed over a spatial extent of 36 km by 36 km. The network is separated into multiple sub-networks, each having up to 30 nodes, whose location is selected in part based on maximizing the land cover diversity within the 36 km cell. The network has achieved unprecedented energy efficiency, longevity, and spatial coverage using custom-designed hardware and software protocols. The network architecture utilizes a nested strategy, where a number of end devices (EDs) communicate to a local coordinator (LC) using our recently developed hardware with ultra-efficient circuitry and best-effort-timeslot allocation communication protocol. The LCs in turn communicates with the base station (BS) via text messages and a new compression scheme. The data collected is made publicly available, in real-time, via the project website (soilscape.usc.edu).