Abstract
Accurate wetland mapping is essential for monitoring ecosystem health and guiding conservation planning efforts. This study evaluated the effectiveness of deep learning architectures and high-resolution National Agriculture Imagery Program (NAIP) imagery for classifying diverse vegetation cover across five public land holdings in three states with significant seasonally flooded, emergent wetland habitats in the southeastern United States. We built and tested deep learning models in ArcGIS Pro using three architectural encoders (U-Net, PSPNet, and DeepLabV3) with different input combinations, including NAIP natural color (RGB), Near-Infrared bands (NIR), Normalized Difference Vegetation Index (NDVI), Digital Elevation Models (DEM), and Texture metrics derived from NAIP imagery using Gray Level Cooccurrence Matrix (GLCM) features, resulting in eight (8) input combinations and three (3) rotational augmentations (0°, 30°, and 60°). Results revealed differences in classification accuracy ranging from 71 to 90% across sites, with Kappa scores from 0.54 to 0.81. We also found differences in model performance under alternating conditions and in the impact of combinations on performance metrics and computational time. Classification of the forested and scrub-shrub classes was consistent with F1 scores above 0.90 at several sites. However, distinguishing open water, aquatic beds, and emergent marshes became more difficult for some model-training architectures. At Mingo National Wildlife Refuge, clear spectral signatures and high-quality training data yielded the highest accuracy, thereby improving classification performance. In contrast, classification accuracy was lower for Red Slough and Choctaw East Wildlife Management Areas. Across all scenarios, the U-Net and DeepLabV3 models consistently outperformed the PSPNet model, achieving mean F1 scores close to 0.91, whereas PSPNet underperformed, with several configurations falling below 0.80. These results explain that while deep learning methods are effective and adaptable across different wetland types, their success depends on the ecological diversity of these environments. Overall, this research highlights the potential of deep learning for wetland monitoring and its crucial role in supporting conservation and management efforts.
Similar content being viewed by others
Resolving inherent constraints in eutrophication monitoring of small lakes using multi-source satellites and machine learning
Deep learning framework for mapping nitrate pollution in coastal aquifers under land use pressure
Agricultural sustainability monitoring in arid regions using hybrid deep learning and Landsat 8 imagery in Najran City, Saudi Arabia
Acknowledgements
The authors thank the University of Arkansas at Monticello, College of Forestry, Agriculture and Natural Resources. Support for this research was provided by the Arkansas Forest Resources Center. We are grateful to the Lower Mississippi Valley Joint Venture and the American Bird Conservancy for their professional collaboration. Special gratitude is extended to the National Wildlife Refuge manager who provided expert verification of habitat classes and essential site-specific information. We also appreciate the cooperation of the Oklahoma Department of Wildlife Conservation, the United States Forest Service, and the Natural Resources Conservation Service at the Red Slough Wildlife Management Area. Finally, we acknowledge the United States Fish and Wildlife Service for their assistance regarding the Mingo and Wapanocca National Wildlife Refuges.
Permission to conduct research at study sites
This research was conducted as part of the funded project “Applications in Remote Sensing to Assess the Quality and Quantity of Emergent Marsh for Waterfowl and Other Wetland Bird Species in the Arkansas Delta”, supported by the Lower Mississippi Valley Joint Venture (LMVJV; Grant No. F23AC02581‑00‑UADA‑AES‑UAMF‑USFWS). All research activities were carried out with the authorization and approval of the relevant authorities and collaborating organizations associated with the project. Permission to access and conduct research at the experimental sites was obtained prior to data collection, and all field and data acquisition activities were performed in accordance with institutional policies, funding agency requirements, and applicable regulations, with the knowledge and approval of site managers and project partners.
Funding
Financial support for this project was provided by the University of Arkansas Agricultural Experiment Station, the Lower Mississippi Valley Joint Venture & American Bird Conservancy, and the Snowden Waterfowl Center of Excellence.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Reprints and permissions
About this article
Cite this article
Ogwo, O.C., Zurqani, H.A., Osborne, D.C. et al. Deep learning for wetland vegetation mapping in the southeastern United States: evaluating site-specific accuracy and classification challenges.
Sci Rep (2026). https://doi.org/10.1038/s41598-026-57454-w
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-026-57454-w
Keywords
- Wetland classification
- Deep learning
- Vegetation mapping
- Southeastern United States
- Remote sensing
- Spectral heterogeneity
Source: Ecology - nature.com
