Drainage ditch network extraction from lidar data using deep convolutional neural networks in a low relief landscape

Authors: DU, L. - University Of Maryland; McCarty, Gregory; LI, X. - Beijing Normal University; ZHANG, X. - Manchester Metropolitan University; RABENHORST, M. - University Of Maryland; LANG, M. - Us Fish And Wildlife Service; ZOU, Z. - University Of Maryland; Zhang, Xuesong; HINSON, A. - Orise Fellow

Submitted to: Journal of Hydrology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/18/2023
Publication Date: 12/7/2023
Citation: Du, L., Mccarty, G.W., Li, X., Zhang, X., Rabenhorst, M.C., Lang, M.W., Zou, Z., Zhang, X., Hinson, A.L. 2023. Drainage ditch network extraction from lidar data using deep convolutional neural networks in a low relief landscape . Journal of Hydrology. 628. https://doi.org/10.1016/j.jhydrol.2023.130591.
DOI: https://doi.org/10.1016/j.jhydrol.2023.130591

Interpretive Summary: Man-made drainage ditch networks in agricultural landscapes facilitate rapid movement of agricultural contaminants to the stream network of watersheds. Ability to accurately characterize the ditch network is important for understanding agricultural sources of contaminants, but such networks have been very difficult to map using standard methods. In this study, we developed deep learning artificial intelligence (AI) models using high resolution topographic information from lidar survey data to map the ditch network in a low relief landscape. We found that the deep learning prediction model outperformed other more conventional prediction methods and created well-connected ditch maps that are free from model noise. Ditch maps produced by this state-of-the-art AI mapping approach will permit better modeling of agricultural sources of stream water contaminants and improve the ability to implement conservation practices to mitigate contamination of surface water resources in agricultural watersheds.

Technical Abstract: Drainage networks composed of small, channelized ditches are very common in the eastern United States. These are human-made features commonly constructed for wetland drainage and constitute the headwater portion of permanent hydrographic networks. Accurate information on the drainage ditch location can help define where wetlands have been drained and evaluate impacts of artificial drainage patterns on hydrologic changes. Traditional water channel extraction approaches often cannot accurately identify small ditches especially in low-relief agricultural landscapes. In this study, we employed a state-of-the-art deep learning (DL) approach to extract drainage ditches using light detection and ranging (lidar) data in a low-relief agricultural landscape within the Delmarva area. First, we adopted a deep convolutional neural network based on U-Net architecture to classify ditches from different combinations of aerial optical and lidar derived (i.e., topographic and return intensity) features. The classification results were compared with a typical pixel-oriented machine learning classifier, random forest (RF). Next, we improved the connectivity of ditch networks through a minimum-cost approach and connected them with natural drainage networks. Finally, we evaluated the connected drainage networks against flowlines derived by traditional flow routing methods and from the U.S. Geological Survey National Hydrography Dataset High Resolution data at 1:24,000 scale. Our results show that the DL model significantly outperformed the RF model, and the lidar derived topographic features were the most important input for ditch classification. The connected drainage networks extracted with DL exhibit pronouncedly higher precision (0.85) and recall (0.86) and a higher positional accuracy (within one pixel) than other flowline products. Overall, this study demonstrates the utility of DL approaches for automated extraction of ditches and the important contribution of high-resolution lidar data for operational drainage network mapping at local and regional scales.