Berkeley Lab scientists are studying a watershed site near the Upper Colorado River where early snowmelt could impact how much water is available to millions of Americans who need it downstream. For now, like many other researchers studying environmental processes, this team lacks access to timely detailed data about when changes like snowmelt or rainfall might occur at remote research sites. This can make it difficult to plan on-the-ground monitoring.
Yuxin Wu, EESA research scientist, is leading an effort to establish integrated networks of multiple sensors capable of remote telemetry to help bring the data environmental scientists need to their desktops on a more timely basis. These solutions will advance new approaches to measure and connect sensors in soils and plants, and use machine learning and advanced data analytics for interpretation. They are being designed not just to speed up the process of receiving data obtained from autonomously derived field measurements, but to ‘watch’ interactions between above- and below-ground processes as they occur. The approach of using above-and-below ground data to observe these interactions has recently been described by Baptiste Dafflon.
To improve understanding of soil-plant interactions, a SMART (Sensors at Mesoscale for Autonomous Remote Telemetry) Soils Testbed is being developed near Building 83 at Berkeley Lab. SMART Soils will be part of EcoSense, which is related to the lab’s Microbes to Biomes initiative.

Berkeley Lab researchers are looking for ways to improve access to and interpret data derived from autonomous advanced sensors and sensor networks to quantify biological-environmental feedbacks across scales, such as from individual plants to entire watersheds or from nanometers to meters.
Lab researchers are currently establishing advanced sensors and sensor networks to quantify biological-environmental feedbacks across scales, such as from individual plants to entire watersheds or from nanometers to meters. Under EcoSENSE, they will also develop ways to synchronize and provide virtual connections between laboratory, controlled mesoscale and field experiments. This will help research scientists gather insights about ecosystem function, translations from microbe to biome and the potential to control biome behavior.
With the SMART Soils Testbed, Wu and other scientists at Berkeley Lab will replicate a hydrogeochemical setting similar to what you’d find in terrestrial ecosystems, which could include a mountainous watershed site and agricultural field, or a contaminated region. Plans are to establish a mobile ecosystem on wheels that can be relocated if necessary. Tilted to represent the hydrological gradient, the testbed will allow Wu and others to measure the quantity and rate of evapotranspiration from plants using multiple sensors, such as load cells. Another objective is for SMART Soils to evaluate how capable sensors are at measuring not just physical properties, but also chemical and biological properties, such as the nitrate or phosphate concentration of soils.
Most important is for the SMART Soils testbed to help measure how well integrated networks of multiple autonomous sensors can gather diverse data and quantify how these various data relate to one another and enable real-time analysis by environmental scientists.
“We want to leverage advanced data analytics so scientists don’t have to look at environmental data in isolation,” Wu says. “Autonomous and telemetric sensors can measure the soil’s moisture or nutrient concentration, for example, and we can design the analytics so that a researcher can evaluate that information to understand how those measurements relate to one another. The telemetry capability is critical in providing real-time measurements that can trigger field-sampling efforts, which often times happen in very remote locations.”
Yuxin Wu is leading a workshop at Berkeley Lab related to this subject on December 1. Please contact him if you are interested in attending.