Developing a smart sensor network for soil moisture monitoring in forests

Forests play a crucial role in regulating the water content of soils and thus influence runoff formation, but also the susceptibility to drought or forest fires. However, the extent to which forests influence soil moisture is difficult to quantify and depends on several parameters, such as precipitation intensity and duration, and terrain or soil properties. To capture the temporal and spatial variability of soil moisture in forests, large-scale and long-term measurements are necessary. Currently, such measurements are relatively expensive and complex and thus generally lacking or restricted to agricultural areas.  

Our current work focuses on the development of a low-cost soil moisture sensor that uses off the shelf parts and can be deployed at scale to provide continuous long-term measurements. To increase adoption and ensure the digital sustainability of our concept, the project will be released open source to the general public.  

The sensor design is based around an ESP32 microcontroller to manage measurements with capacitive soil moisture sensors. For communication, we leverage the LoRa protocol and use infrastructure provided by the Things Network (TTN). Herein, we present the soft- and hardware architecture of a sensor prototype and results obtained from a proof-of-concept deployment. In addition, we discuss the calibration procedure and evaluation of capacitive soil moisture sensors (in comparison to time-domain reflectometry (TDR) sensors). Finally, we provide an outlook on future developments of our measurement system. The final goal of this project is to deploy sensors in several areas of interest that will allow for gathering data for a better understanding of the interaction of forests and soil moisture content.