{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,6]],"date-time":"2026-04-06T14:32:21Z","timestamp":1775485941960,"version":"3.50.1"},"reference-count":60,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,5,13]],"date-time":"2025-05-13T00:00:00Z","timestamp":1747094400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:p>High-resolution digital elevation models (HRDEMs) from LiDAR and InSAR technologies have significantly improved the accuracies of mapping hydrographic features such as river boundaries, streamlines, and waterbodies over large areas. However, drainage crossings that facilitate the passage of drainage flows beneath roads are not often represented in HRDEMs, resulting in erratic or distorted hydrographic features. At present, drainage crossing datasets are largely missing or available with variable quality. While previous studies have investigated basic convolutional neural network (CNN) models for drainage crossing characterization, it remains unclear if advanced deep learning models will improve the accuracy of drainage crossing classification. Although HRDEM-derived geomorphological features have been identified to enhance feature extraction in other hydrography applications, the contributions of these features to drainage crossing image classification have yet to be sufficiently investigated. This study develops advanced CNN models, EfficientNetV2, using four co-registered 1-meter resolution geomorphological data layers derived from HRDEMs for drainage crossing classification. These layers include positive openness (POS), geometric curvature, and two topographic position index (TPI) layers utilizing 3 \u00d7 3 and 21 \u00d7 21 cell windows. The findings reveal that the advanced CNN models with HRDEM, TPI (21 \u00d7 21), and a combination of HRDEM, POS, and TPI (21 \u00d7 21) improve classification accuracy in comparison to the baseline model by 3.39, 4.27, and 4.93%, respectively. The study culminates in explainable artificial intelligence (XAI) for evaluating those most critical image segments responsible for characterizing drainage crossings.<\/jats:p>","DOI":"10.3389\/frai.2025.1561281","type":"journal-article","created":{"date-parts":[[2025,5,13]],"date-time":"2025-05-13T05:31:07Z","timestamp":1747114267000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":2,"title":["Deep learning classification of drainage crossings based on high-resolution DEM-derived geomorphological information"],"prefix":"10.3389","volume":"8","author":[{"given":"Michael","family":"Edidem","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Bill","family":"Xu","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ruopu","family":"Li","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Di","family":"Wu","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Banafsheh","family":"Rekabdar","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Guangxing","family":"Wang","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1965","published-online":{"date-parts":[[2025,5,13]]},"reference":[{"key":"ref1","doi-asserted-by":"publisher","first-page":"171","DOI":"10.5552\/crojfe.2023.1701","article-title":"Use of UAV data and HEC-RAS model for dimensioning of hydraulic structures on Forest roads","volume":"44","author":"A\u00e7\u0131l","year":"2023","journal-title":"Croat. J. For. Eng."},{"key":"ref2","volume-title":"National Water Center Innovators Program Summer Institute Report 2018","author":"Aristizabal","year":"2018"},{"key":"ref3","doi-asserted-by":"publisher","first-page":"121","DOI":"10.1080\/24705357.2022.2040388","article-title":"Remote sensing framework details Riverscape connectivity fragmentation and fish Passability in a forested landscape","volume":"8","author":"Arsenault","year":"2023","journal-title":"J. Ecohydraul."},{"key":"ref4","doi-asserted-by":"publisher","first-page":"401","DOI":"10.1559\/152304005775194692","article-title":"Lidar elevation data for surface hydrologic modeling: resolution and representation issues","volume":"32","author":"Barber","year":"2005","journal-title":"Cartogr. Geogr. Inf. Sci."},{"key":"ref5","doi-asserted-by":"publisher","first-page":"234","DOI":"10.3390\/rs14010234","article-title":"Stream boundary detection of a hyper-arid, polar region using a U-net architecture: Taylor Valley, Antarctica","volume":"14","author":"Barlow","year":"2022","journal-title":"Remote Sens."},{"key":"ref6","doi-asserted-by":"publisher","first-page":"2596","DOI":"10.1111\/tgis.12832","article-title":"Assessing the impacts of anthropogenic drainage structures on hydrologic connectivity using high-resolution digital elevation models","volume":"25","author":"Bhadra","year":"2021","journal-title":"Trans. GIS"},{"key":"ref7","doi-asserted-by":"publisher","first-page":"541","DOI":"10.1002\/hyp.9224","article-title":"Drainage network detection and assessment of network storage capacity in agrarian landscape","volume":"27","author":"Cazorzi","year":"2013","journal-title":"Hydrol. Process."},{"key":"ref8","doi-asserted-by":"publisher","first-page":"20160706","DOI":"10.1098\/rspa.2016.0706","article-title":"A restatement of the natural science evidence concerning catchment-based \u2018natural\u2019 flood management in the UK","volume":"473","author":"Dadson","year":"2017","journal-title":"Proc. R. Soc. Math. Phys. Eng. Sci."},{"key":"ref9","doi-asserted-by":"publisher","first-page":"39","DOI":"10.1016\/j.geomorph.2012.12.015","article-title":"Application of the topographic position index to heterogeneous landscapes","volume":"186","author":"De Reu","year":"2013","journal-title":"Geomorphology"},{"key":"ref10","doi-asserted-by":"publisher","first-page":"843","DOI":"10.1002\/jpln.200900094","article-title":"A multiscale soil\u2013landform relationship in the glacial-drift area based on digital terrain analysis and soil attributes","volume":"173","author":"Deumlich","year":"2010","journal-title":"J. Plant Nutr. Soil Sci."},{"key":"ref11","doi-asserted-by":"publisher","first-page":"6427","DOI":"10.3390\/rs5126427","article-title":"Openness as visualization technique for interpretative mapping of airborne Lidar derived digital terrain models","volume":"5","author":"Doneus","year":"2013","journal-title":"Remote Sens."},{"key":"ref12","doi-asserted-by":"publisher","first-page":"1827","DOI":"10.1002\/hyp.5964","article-title":"Incorporating ancillary data to refine Anthropogenically modified overland flow paths","volume":"20","author":"Duke","year":"2006","journal-title":"Hydrol. Process."},{"key":"ref13","doi-asserted-by":"publisher","first-page":"106338","DOI":"10.1016\/j.envsoft.2025.106338","article-title":"GeoAI-based drainage crossing detection for elevation-derived hydrographic mapping","volume":"186","author":"Edidem","year":"2025","journal-title":"Environ. Model Softw."},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.2011.07584","article-title":"Pix2Streams: dynamic hydrology maps from satellite-LiDAR fusion","author":"Garcia","year":"2020","journal-title":"arXiv"},{"key":"ref15","volume-title":"Automation of DEM cutting for hydrologic\/hydraulic modeling","author":"Gelder","year":"2015"},{"key":"ref16","doi-asserted-by":"publisher","first-page":"175","DOI":"10.1126\/science.aaa5931","article-title":"Hydrologic connectivity constrains partitioning of global terrestrial water fluxes","volume":"349","author":"Good","year":"2015","journal-title":"Science"},{"key":"ref17","doi-asserted-by":"publisher","first-page":"e2015759118","DOI":"10.1073\/pnas.2015759118","article-title":"Replication across space and time must be weak in the social and environmental sciences","volume":"118","author":"Goodchild","year":"2021","journal-title":"Proc. Natl. Acad. Sci."},{"key":"ref52","first-page":"198","article-title":"Stream guiding algorithm for deriving flow direction from dem and location of main streams","volume-title":"cold region hydrology in a changing climate: proceedings of symposium h02 held during IUGG2011","author":"Hao","year":"2011"},{"key":"ref18","doi-asserted-by":"publisher","first-page":"21101","DOI":"10.1007\/s00521-022-07593-8","article-title":"Prediction of hydraulic blockage at culverts from a single image using deep learning","volume":"34","author":"Iqbal","year":"2022","journal-title":"Neural Comput. & Applic."},{"key":"ref19","first-page":"1","article-title":"Flood detection from satellite images using self-attention empowered EfficientNetV2 mechanism","author":"Jeba","year":"2023"},{"key":"ref20","doi-asserted-by":"publisher","first-page":"427","DOI":"10.1080\/13658816.2020.1802032","article-title":"Reproducibility and replicability: opportunities and challenges for geospatial research","volume":"35","author":"Kedron","year":"2021","journal-title":"Int. J. Geogr. Inf. Sci."},{"key":"ref21","first-page":"2668","article-title":"Interpretability beyond feature attribution: quantitative testing with concept activation vectors (TCAV)","volume-title":"International conference on machine learning","author":"Kim","year":"2017"},{"key":"ref22","doi-asserted-by":"publisher","first-page":"248","DOI":"10.1080\/07011784.2022.2160660","article-title":"Performance of automated geoprocessing methods for culvert detection in remote Forest environments","volume":"48","author":"Lessard","year":"2023","journal-title":"Can. Water Resour. J."},{"key":"ref23","doi-asserted-by":"publisher","first-page":"101845","DOI":"10.1016\/j.compenvurbsys.2022.101845","article-title":"Extracting spatial effects from machine learning model using local interpretation method: an example of SHAP and XGBoost","volume":"96","author":"Li","year":"2022","journal-title":"Comput. Environ. Urban. Syst."},{"key":"ref24","doi-asserted-by":"crossref","first-page":"151","DOI":"10.1007\/978-981-19-3816-0_17","article-title":"GeoAI and the future of spatial analytics","volume-title":"New thinking in GIScience","author":"Li","year":"2022"},{"key":"ref25","doi-asserted-by":"publisher","first-page":"1136","DOI":"10.3390\/ijgi2041136","article-title":"Drainage structure datasets and effects on LiDAR-derived surface flow modeling","volume":"2","author":"Li","year":"2013","journal-title":"ISPRS Int. J. Geo Inf."},{"key":"ref26","doi-asserted-by":"publisher","first-page":"107045","DOI":"10.1016\/j.geomorph.2020.107045","article-title":"Deep learning-based approach for landform classification from integrated data sources of digital elevation model and imagery","volume":"354","author":"Li","year":"2020","journal-title":"Geomorphology"},{"key":"ref27","doi-asserted-by":"publisher","first-page":"4660","DOI":"10.1002\/hyp.11385","article-title":"Evaluating preprocessing methods of digital elevation models for hydrological modelling","volume":"31","author":"Lidberg","year":"2017","journal-title":"Hydrol. Process."},{"key":"ref28","doi-asserted-by":"publisher","first-page":"846","DOI":"10.1002\/hyp.10648","article-title":"Efficient hybrid breaching-filling sink removal methods for flow path enforcement in digital elevation models: efficient hybrid sink removal methods for flow path enforcement","volume":"30","author":"Lindsay","year":"2016","journal-title":"Hydrol. Process."},{"key":"ref29","doi-asserted-by":"publisher","first-page":"397","DOI":"10.1080\/13658816.2014.975715","article-title":"Modelling surface drainage patterns in altered landscapes using LiDAR","volume":"29","author":"Lindsay","year":"2015","journal-title":"Int. J. Geogr. Inf. Sci."},{"key":"ref31","doi-asserted-by":"publisher","first-page":"2011WR010958","DOI":"10.1029\/2011WR010958","article-title":"Automatic geomorphic feature extraction from Lidar in flat and engineered landscapes","volume":"48","author":"Passalacqua","year":"2012","journal-title":"Water Resour. Res."},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.6028\/nist.ir.8312","author":"Phillips","year":"2020","journal-title":"Four principles of explainable artificial intelligence"},{"key":"ref33","doi-asserted-by":"publisher","first-page":"1187","DOI":"10.1002\/hyp.7582","article-title":"Suitability of LiDAR point density and derived landform curvature maps for channel network extraction","volume":"24","author":"Pirotti","year":"2010","journal-title":"Hydrol. Process."},{"key":"ref34","doi-asserted-by":"publisher","first-page":"90","DOI":"10.2112\/SI76-009","article-title":"Hydrologic connectivity: quantitative assessments of hydrologic-enforced drainage structures in an elevation model","volume":"76","author":"Poppenga","year":"2016","journal-title":"J. Coast. Res."},{"key":"ref35","article-title":"Using selective drainage methods to extract continuous surface flow from 1-meter Lidar-derived digital elevation data","volume-title":"Sci. Rep.","author":"Poppenga","year":"2010"},{"key":"ref36","doi-asserted-by":"publisher","first-page":"981","DOI":"10.1890\/1051-0761(2001)011[0981:hcatmo]2.0.CO;2","article-title":"Hydrologic connectivity and the Management of Biological Reserves: a global perspective","volume":"11","author":"Pringle","year":"2001","journal-title":"Ecol. Appl."},{"key":"ref37","first-page":"87","article-title":"Chapter 4 preparation of DEMs for Geomorphometric analysis","volume-title":"Developments in soil science","author":"Reuter","year":"2009"},{"key":"ref38","doi-asserted-by":"publisher","first-page":"58","DOI":"10.1016\/j.envsoft.2016.04.026","article-title":"GeoNet: an open-source software for the automatic and objective extraction of channel heads, channel network, and channel morphology from high resolution topography data","volume":"83","author":"Sangireddy","year":"2016","journal-title":"Environ. Model Softw."},{"key":"ref39","doi-asserted-by":"publisher","first-page":"60","DOI":"10.1016\/j.jaridenv.2017.05.008","article-title":"Spatial and temporal variability in the effects of wildfire and drought on thermal habitat for a desert trout","volume":"145","author":"Schultz","year":"2017","journal-title":"J. Arid Environ."},{"key":"ref40","doi-asserted-by":"publisher","first-page":"1576","DOI":"10.1002\/esp.3452","article-title":"Flow network derivation from a high resolution DEM in a low relief, agrarian landscape","volume":"38","author":"Schwanghart","year":"2013","journal-title":"Earth Surf. Process. Landf."},{"key":"ref41","doi-asserted-by":"publisher","first-page":"336","DOI":"10.1007\/s11263-019-01228-7","article-title":"Grad-CAM: visual explanations from deep networks via gradient-based localization","volume":"128","author":"Selvaraju","year":"2020","journal-title":"Int. J. Comput. Vis."},{"key":"ref42","doi-asserted-by":"publisher","first-page":"97","DOI":"10.3390\/ijgi5060097","article-title":"Parallel Landscape Driven Data Reduction & Spatial Interpolation Algorithm for big LiDAR data","volume":"5","author":"Sharma","year":"2016","journal-title":"ISPRS Int. J. Geo Inf."},{"key":"ref43","doi-asserted-by":"publisher","first-page":"7","DOI":"10.1016\/j.envsoft.2013.04.003","article-title":"Evaluation of a surface hydrological connectivity index in agricultural catchments","volume":"47","author":"Shore","year":"2013","journal-title":"Environ. Model Softw."},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.3133\/fs20093054","article-title":"The National Map - Hydrography","author":"Simley","year":"2009","journal-title":"Fact Sheet"},{"key":"ref45","doi-asserted-by":"publisher","first-page":"2046","DOI":"10.1002\/hyp.9727","article-title":"High-resolution topography and anthropogenic feature extraction: testing Geomorphometric parameters in floodplains","volume":"28","author":"Sofia","year":"2014","journal-title":"Hydrol. Process."},{"key":"ref46","doi-asserted-by":"publisher","first-page":"2368","DOI":"10.3390\/rs13122368","article-title":"Extensibility of U-net neural network model for hydrographic feature extraction and implications for hydrologic modeling","volume":"13","author":"Stanislawski","year":"2021","journal-title":"Remote Sens."},{"key":"ref47","doi-asserted-by":"publisher","first-page":"319","DOI":"10.1080\/15230406.2017.1337524","article-title":"An open source high-performance solution to extract surface water drainage networks from diverse terrain conditions","volume":"45","author":"Stanislawski","year":"2018","journal-title":"Cartogr. Geogr. Inf. Sci."},{"key":"ref48","unstructured":"Stutheit\n              R. G.\n            \n            \n              Gilbert\n              M. C.\n            \n            \n              Whited\n              P. M.\n            \n            \n              Lawrence\n              K. L.\n            \n          \n          Vicksburg, MS\n          U.S. Army Engineer Research and Development Center\n          A regional guidebook for applying the hydrogeomorphic approach to assessing wetland functions of rainwater basin depressional wetlands in nebraska\n          \n          2004"},{"key":"ref49","first-page":"137","article-title":"Classification and feature extraction for hydraulic structures data using advanced CNN architectures","author":"Talafha","year":"2021"},{"key":"ref50","volume-title":"EfficientNetV2: Smaller models and faster training","author":"Tan","year":"2021"},{"key":"ref51","doi-asserted-by":"publisher","first-page":"1308","DOI":"10.3390\/rs13071308","article-title":"Automated mapping of transportation embankments in fine-resolution LiDAR DEMs","volume":"13","author":"Van Nieuwenhuizen","year":"2021","journal-title":"Remote Sens."},{"key":"ref53","doi-asserted-by":"publisher","first-page":"193","DOI":"10.1080\/13658810500433453","article-title":"An efficient method for identifying and filling surface depressions in digital elevation models for hydrologic analysis and modelling","volume":"20","author":"Wang","year":"2006","journal-title":"Int. J. Geogr. Inf. Sci."},{"key":"ref54","doi-asserted-by":"publisher","first-page":"83","DOI":"10.1080\/19475683.2019.1604571","article-title":"Review on algorithms of dealing with depressions in grid Dem","volume":"25","author":"Wang","year":"2019","journal-title":"Ann. GIS"},{"key":"ref55","unstructured":"Wood\n              J\n            \n          \n          University of Leicester\n          The geomorphological characterisation of digital elevation models\n          \n          1996"},{"key":"ref56","doi-asserted-by":"publisher","first-page":"1253","DOI":"10.1111\/1752-1688.13239","article-title":"Enhancing hydrologic Lidar digital elevation models: bridging hydrographic gaps at fine scales","volume":"60","author":"Wu","year":"2024","journal-title":"J. Am. Water Resour. Assoc."},{"key":"ref57","doi-asserted-by":"publisher","first-page":"2230706","DOI":"10.1080\/15481603.2023.2230706","article-title":"Classification of drainage crossings on high-resolution digital elevation models: a deep learning approach","volume":"60","author":"Wu","year":"2023","journal-title":"GISci. Remote Sensing"},{"key":"ref58","doi-asserted-by":"publisher","first-page":"423","DOI":"10.1016\/j.isprsjprs.2018.08.005","article-title":"A 3D convolutional neural network method for land cover classification using LiDAR and multi-temporal Landsat imagery","volume":"144","author":"Xu","year":"2018","journal-title":"ISPRS J. Photogramm. Remote Sens."},{"key":"ref59","doi-asserted-by":"publisher","first-page":"104992","DOI":"10.1016\/j.envsoft.2021.104992","article-title":"An attention U-net model for detection of fine-scale hydrologic streamlines","volume":"140","author":"Xu","year":"2021","journal-title":"Environ. Model Softw."},{"key":"ref60","doi-asserted-by":"crossref","DOI":"10.1109\/ICSP54964.2022.9778437","article-title":"Research on remote sensing image classification algorithm based on EfficientNet","author":"Yin","year":"2022"},{"key":"ref61","first-page":"2921","article-title":"Learning deep features for discriminative localization","author":"Zhou","year":"2016"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2025.1561281\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,5,13]],"date-time":"2025-05-13T05:31:12Z","timestamp":1747114272000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2025.1561281\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,5,13]]},"references-count":60,"alternative-id":["10.3389\/frai.2025.1561281"],"URL":"https:\/\/doi.org\/10.3389\/frai.2025.1561281","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,5,13]]},"article-number":"1561281"}}