Peatland Groundwater Level (GWL) Prediction Using Machine Learning through an IoT System

Peatlands play a crucial role in carbon sequestration and biodiversity conservation, yet they are increasingly threatened by unsustainable land-use changes, especially in Southeast Asia.A key aspect of preserving peatlands lies in maintaining appropriate groundwater levels (GWL), which, if lowered, can lead to severe consequences such as peat degradation, increased fire risk, and greenhouse gas emissions.Traditional GWL monitoring methods are labor-intensive and infrequent, resulting in inadequate data for predictive modeling.This study proposes an integrated approach that combines Internet of Things (IoT) technology with machine learning to monitor and predict GWL in real time, enabling more effective peatland management.The IoT system utilizes a network of sensors deployed in peatland areas to collect real-time environmental data, including soil moisture, temperature, rainfall, and water table depth.This data is transmitted to a cloud-based platform where preprocessing and storage take place.The system then applies machine learning modelssuch as Random Forest, Support Vector Machines, and Long Short-Term Memory (LSTM) neural networks-to analyze temporal and spatial patterns in the data to predict future GWL levels.This hybrid approach offers a robust, scalable, and automated solution for early detection of critical GWL thresholds, which can inform timely conservation actions, including rewetting programs.Experimental results indicate that the LSTM model, in particular, demonstrates superior performance in capturing non-linear temporal trends, achieving over 90% accuracy in predictive tasks.The integration of IoT and ML significantly reduces manual labor while enhancing decision-making capabilities.This research contributes a novel, real-time GWL monitoring system, advancing the tools available for environmental management.The proposed model not only enables predictive maintenance of peatland hydrology but also supports broader climate change mitigation efforts by reducing peatland degradation and its associated emissions.