Mapping artificial drains in Peatlands – A national scale assessment of Irish raised bogs using sub-meter aerial imagery and deep learning methods

This paper presents a novel approach to mapping artificial drainage networks in Irish raised bogs using very high-resolution aerial imagery and deep learning methods. The study developed semantic segmentation models to automatically detect and map drainage features across national-scale peatland systems.

Key Contributions

• Developed deep learning workflows in R using Keras and TensorFlow for automated peatland drainage mapping
• Processed 2 TB of aerial imagery using optimized GPU-based training pipelines
• Achieved national-scale assessment of drainage density and intensity across Irish raised bogs
• Delivered spatial metrics to inform peatland conservation policy and land use planning

This research provides crucial insights into peatland degradation patterns and supports evidence-based conservation strategies for these critical carbon storage ecosystems.

Recommended citation: Habib, W., Cresson, R., McGuinness, K., & Connolly, J. (2024). "Mapping artificial drains in Peatlands – A national scale assessment of Irish raised bogs using sub-meter aerial imagery and deep learning methods." Remote Sensing in Ecology and Conservation.
Download Paper