The Internet of Trees: Towards Digital Forest Environments using LoRaWAN

Abstract

The need for high resolution and continuous environmental data, particularly for advancing Natural Flood Management through hydrological monitoring and assessing forest health via Volatile Organic Compound (VOC) detection, demands innovative and autonomous sensing solutions. Traditional data collection methods in these often inaccessible and geographically challenging areas are hampered by significant logistical and financial constraints. The research presented in this thesis was motivated by the opportunity to leverage the emerging capabilities of Internet of Things (IoT) technology, specifically the LoRaWAN Low Power Wide Area Network protocol, to establish cost-effective and scalable Wireless Sensor Networks (WSNs) for long-term environmental monitoring. Focussing on a remote rural catchment, the Pennal catchment in Wales, this thesis aims to move beyond preliminary investigations of LoRaWAN's suitability by undertaking a comprehensive, extended real-world deployment. A key driver is the necessity to integrate specialised environmental sensors within a low-power framework, thereby enabling enhanced ecological understanding and contributing to the growing field of digital natural environments and more informed environmental management strategies. The work presented in this thesis describes the successful design, implementation and rigorous evaluation of a LoRaWAN-based WSN tailored for remote environmental monitoring. A primary achievement was the development and 26-month field deployment of a hydrological monitoring system, demonstrating significant engineered improvements in data acquisition and overall network reliability achieved through iterative optimisation of the hardware, software and power management subsystem. The thesis provides a cost analysis of WSN components, highlighting the substantial influence of sensor selection on the economic viability of such deployments and offering practical recommendations. Furthermore, a highly reliable, renewable energy-powered LoRaWAN gateway was engineered and deployed, achieving over 99\% uptime, showcasing a robust and economically viable solution for remote network infrastructure. A computationally efficient embedded State of Charge (SoC) estimator was developed and integrated to enhance the energy management and availability probability of gateways. This thesis also pioneered the integration of Photoionisation Detectors (PIDs) into a LoRaWAN architecture for VOC monitoring, involving extensive calibrations that, while revealing the inherent complexities in achieving high-accuracy ambient measurements, established a critical foundation for future advancements in low-power autonomous VOC sensing. Ultimately, this thesis provides a comprehensive case study of a real-world deployment of WSN in a challenging remote environment, detailing the practical obstacles encountered and the innovative solutions implemented, offering valuable lessons and advancements for the broader environmental monitoring community.

Date of Award	2025
Original language	English
Supervisor	Iestyn Pierce (Supervisor)