{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T02:47:32Z","timestamp":1760150852595,"version":"build-2065373602"},"reference-count":40,"publisher":"MDPI AG","issue":"3","license":[{"start":{"date-parts":[[2022,1,29]],"date-time":"2022-01-29T00:00:00Z","timestamp":1643414400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Remote Sensing"],"abstract":"<jats:p>With the increasing availability of SAR imagery in recent years, more research is being conducted using deep learning (DL) for the classification of ice and open water; however, ice and open water classification using conventional DL methods such as convolutional neural networks (CNNs) is not yet accurate enough to replace manual analysis for operational ice chart mapping. Understanding the uncertainties associated with CNN model predictions can help to quantify errors and, therefore, guide efforts on potential enhancements using more\u2013advanced DL models and\/or synergistic approaches. This paper evaluates an approach for estimating the aleatoric uncertainty [a measure used to identify the noise inherent in data] of CNN probabilities to map ice and open water with a custom loss function applied to RADARSAT\u20132 HH and HV observations. The images were acquired during the 2014 ice season of Lake Erie and Lake Ontario, two of the five Laurentian Great Lakes of North America. Operational image analysis charts from the Canadian Ice Service (CIS), which are based on visual interpretation of SAR imagery, are used to provide training and testing labels for the CNN model and to evaluate the accuracy of the model predictions. Bathymetry, as a variable that has an impact on the ice regime of lakes, was also incorporated during model training in supplementary experiments. Adding aleatoric loss and bathymetry information improved the accuracy of mapping water and ice. Results are evaluated quantitatively (accuracy metrics) and qualitatively (visual comparisons). Ice and open water scores were improved in some sections of the lakes by using aleatoric loss and including bathymetry. In Lake Erie, the ice score was improved by \u223c2 on average in the shallow near\u2013shore zone as a result of better mapping of dark ice (low backscatter) in the western basin. As for Lake Ontario, the open water score was improved by \u223c6 on average in the deepest profundal off\u2013shore zone.<\/jats:p>","DOI":"10.3390\/rs14030644","type":"journal-article","created":{"date-parts":[[2022,1,29]],"date-time":"2022-01-29T01:43:27Z","timestamp":1643420607000},"page":"644","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":7,"title":["Incorporating Aleatoric Uncertainties in Lake Ice Mapping Using RADARSAT\u20132 SAR Images and CNNs"],"prefix":"10.3390","volume":"14","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-2397-2040","authenticated-orcid":false,"given":"Nastaran","family":"Saberi","sequence":"first","affiliation":[{"name":"Department of Geography and Environmental Management, University of Waterloo, Waterloo, ON N2L 3G1, Canada"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Katharine Andrea","family":"Scott","sequence":"additional","affiliation":[{"name":"Department of Systems Design Engineering, University of Waterloo, Waterloo, ON N2L 3G1, Canada"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1044-5850","authenticated-orcid":false,"given":"Claude","family":"Duguay","sequence":"additional","affiliation":[{"name":"Department of Geography and Environmental Management, University of Waterloo, Waterloo, ON N2L 3G1, Canada"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2022,1,29]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"418","DOI":"10.1038\/nature20584","article-title":"High-resolution mapping of global surface water and its long-term changes","volume":"540","author":"Pekel","year":"2016","journal-title":"Nature"},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"781","DOI":"10.1002\/hyp.6131","article-title":"Recent trends in Canadian lake ice cover","volume":"20","author":"Duguay","year":"2006","journal-title":"Hydrol. Process. Int. J."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"671","DOI":"10.1177\/0309133310375653","article-title":"The response and role of ice cover in lake-climate interactions","volume":"34","author":"Brown","year":"2010","journal-title":"Prog. Phys. Geogr. Earth Environ."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"3942","DOI":"10.1002\/joc.5546","article-title":"Climatological trends of snowfall over the Laurentian Great Lakes Basin","volume":"38","author":"Duguay","year":"2018","journal-title":"Int. J. Climatol."},{"key":"ref_5","first-page":"218","article-title":"Implementation of the lake parameterisation scheme FLake into the numerical weather prediction model COSMO","volume":"15","author":"Mironov","year":"2010","journal-title":"Boreal Environ. Res."},{"key":"ref_6","first-page":"1747","article-title":"The Effects of Lake Representation on the Regional Hydroclimate in the ECMWF Reanalyses","volume":"149","author":"Minallah","year":"2021","journal-title":"Mon. Weather Rev."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"722","DOI":"10.3394\/0380-1330(2007)33[722:SSRSOG]2.0.CO;2","article-title":"Satellite SAR remote sensing of Great Lakes ice cover, part 1. Ice backscatter signatures at C band","volume":"33","author":"Nghiem","year":"2007","journal-title":"J. Great Lakes Res."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"697","DOI":"10.1007\/s10584-013-0840-2","article-title":"Coasts, water levels, and climate change: A Great Lakes perspective","volume":"120","author":"Gronewold","year":"2013","journal-title":"Clim. Chang."},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"2510","DOI":"10.1002\/lno.10585","article-title":"Grand challenges for research in the Laurentian Great Lakes","volume":"62","author":"Sterner","year":"2017","journal-title":"Limnol. Oceanogr."},{"key":"ref_10","unstructured":"Manice, C. (2005). Manual of Standard Procedures for Observing and Reporting Ice Conditions. Environment Canada."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"528","DOI":"10.1109\/TGRS.2004.839589","article-title":"Unsupervised segmentation of synthetic aperture Radar sea ice imagery using a novel Markov random field model","volume":"43","author":"Deng","year":"2005","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"65","DOI":"10.3189\/2013AoG62A037","article-title":"Observing lake- and river-ice decay with SAR: Advantages and limitations of the unsupervised k-means classification approach","volume":"54","author":"Sobiech","year":"2013","journal-title":"Ann. Glaciol."},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Wang, J., Duguay, C.R., Clausi, D.A., Pinard, V., and Howell, S.E. (2018). Semi-automated classification of Lake Ice Cover using dual polarization RADARSAT-2 imagery. Remote Sens., 10.","DOI":"10.3390\/rs10111727"},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"5529","DOI":"10.1109\/TGRS.2013.2290231","article-title":"Automated Ice\u2013Water Classification Using Dual Polarization SAR Satellite Imagery","volume":"52","author":"Leigh","year":"2014","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_15","doi-asserted-by":"crossref","unstructured":"Hoekstra, M., Jiang, M., Clausi, D.A., and Duguay, C. (2020). Lake ice-water classification of RADARSAT-2 images by integrating IRGS Segmentation with pixel-based random forest labeling. Remote Sens., 12.","DOI":"10.3390\/rs12091425"},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"4397","DOI":"10.1109\/TGRS.2012.2192278","article-title":"Operational SAR sea-ice image classification","volume":"50","author":"Ochilov","year":"2012","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"S391","DOI":"10.5589\/m11-001","article-title":"Monitoring lake ice during spring melt using RADARSAT-2 SAR","volume":"36","author":"Geldsetzer","year":"2010","journal-title":"Can. J. Remote Sens."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"5558","DOI":"10.1109\/TGRS.2013.2290331","article-title":"Baltic Sea Ice Concentration Estimation Based on C-Band Dual-Polarized SAR Data","volume":"52","author":"Karvonen","year":"2014","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"1045","DOI":"10.1109\/TGRS.2009.2031806","article-title":"Mapping of Different Sea Ice Regimes Using Images From Sentinel-1 and ALOS Synthetic Aperture Radar","volume":"48","author":"Dierking","year":"2010","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"33","DOI":"10.5194\/tc-11-33-2017","article-title":"Operational algorithm for ice\u2013water classification on dual-polarized RADARSAT-2 images","volume":"11","author":"Zakhvatkina","year":"2017","journal-title":"Cryosphere"},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"1521","DOI":"10.1016\/j.jglr.2020.08.018","article-title":"Retrieval of ice\/water observations from synthetic aperture radar imagery for use in lake ice data assimilation","volume":"46","author":"Scott","year":"2020","journal-title":"J. Great Lakes Res."},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"409","DOI":"10.5194\/isprs-annals-V-3-2020-409-2020","article-title":"Lake ice detection from sentinel-1 sar with deep learning","volume":"5","author":"Tom","year":"2020","journal-title":"ISPRS Ann. Photogramm. Remote Sens. Spat. Inf. Sci."},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"1763","DOI":"10.1109\/LSP.2017.2758203","article-title":"SAR Image Despeckling Using a Convolutional Neural Network","volume":"24","author":"Wang","year":"2017","journal-title":"IEEE Signal Process. Lett."},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"4843","DOI":"10.1109\/TIP.2017.2725580","article-title":"Going Deeper With Contextual CNN for Hyperspectral Image Classification","volume":"26","author":"Lee","year":"2017","journal-title":"IEEE Trans. Image Process."},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Jiang, T., Cui, Z., Zhou, Z., and Cao, Z. (2018, January 22\u201327). Data Augmentation with Gabor Filter in Deep Convolutional Neural Networks for Sar Target Recognition. Proceedings of the IGARSS 2018\u20142018 IEEE International Geoscience and Remote Sensing Symposium, Valencia, Spain.","DOI":"10.1109\/IGARSS.2018.8518792"},{"key":"ref_26","doi-asserted-by":"crossref","unstructured":"Kang, M., Ji, K., Leng, X., and Lin, Z. (2017). Contextual region-based convolutional neural network with multilayer fusion for SAR ship detection. Remote Sens., 9.","DOI":"10.3390\/rs9080860"},{"key":"ref_27","doi-asserted-by":"crossref","unstructured":"Zhang, W., Tang, P., and Zhao, L. (2019). Remote sensing image scene classification using CNN-CapsNet. Remote Sens., 11.","DOI":"10.3390\/rs11050494"},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1029\/2005JC003384","article-title":"Sea ice remote sensing using AMSR-E 89-GHz channels","volume":"113","author":"Spreen","year":"2008","journal-title":"J. Geophys. Res. Ocean."},{"key":"ref_29","doi-asserted-by":"crossref","unstructured":"Boulze, H., Korosov, A., and Brajard, J. (2020). Classification of sea ice types in sentinel-1 SAR data using convolutional neural networks. Remote Sens., 12.","DOI":"10.3390\/rs12132165"},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"4392","DOI":"10.1109\/TIE.2017.2764844","article-title":"NB-CNN: Deep Learning-Based Crack Detection Using Convolutional Neural Network and Na\u00efve Bayes Data Fusion","volume":"65","author":"Chen","year":"2018","journal-title":"IEEE Trans. Ind. Electron."},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"247","DOI":"10.1109\/TGRS.2020.2992454","article-title":"Evaluation of a Neural Network With Uncertainty for Detection of Ice and Water in SAR Imagery","volume":"59","author":"Asadi","year":"2021","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_32","unstructured":"Kendall, A. (2019). Geometry and Uncertainty in Deep Learning for Computer Vision. [Ph.D. Thesis, University of Cambridge]."},{"key":"ref_33","doi-asserted-by":"crossref","first-page":"457","DOI":"10.1007\/s10994-021-05946-3","article-title":"Aleatoric and epistemic uncertainty in machine learning: An introduction to concepts and methods","volume":"110","author":"Waegeman","year":"2021","journal-title":"Mach. Learn."},{"key":"ref_34","unstructured":"(2020, September 30). Great Lakes Environmental Research Laboratory\u2014Historical Ice Cover, Available online: https:\/\/www.glerl.noaa.gov\/data\/ice\/#historical."},{"key":"ref_35","unstructured":"(2020, September 30). Historical Climate Data. Available online: https:\/\/climate.weather.gc.ca\/historical_data\/search_historic_data_e.html."},{"key":"ref_36","unstructured":"Kendall, A., and Gal, Y. (2017). What uncertainties do we need in Bayesian deep learning for computer vision?. arXiv."},{"key":"ref_37","doi-asserted-by":"crossref","first-page":"1137","DOI":"10.1029\/WR004i005p01137","article-title":"The response of water temperatures to meteorological conditions","volume":"4","author":"Edinger","year":"1968","journal-title":"Water Resour. Res."},{"key":"ref_38","doi-asserted-by":"crossref","first-page":"140","DOI":"10.1061\/(ASCE)0887-381X(2006)20:4(140)","article-title":"Modeling of lake ice characteristics in North America using climate, geography, and lake bathymetry","volume":"20","author":"Williams","year":"2006","journal-title":"J. Cold Reg. Eng."},{"key":"ref_39","doi-asserted-by":"crossref","first-page":"1693","DOI":"10.1139\/cjfas-2017-0242","article-title":"Ecosystem classification and mapping of the laurentian great lakes","volume":"75","author":"Riseng","year":"2018","journal-title":"Can. J. Fish. Aquat. Sci."},{"key":"ref_40","unstructured":"Loshchilov, I., and Hutter, F. (2017, January 24\u201326). SGDR: Stochastic gradient descent with warm restarts. Proceedings of the 5th International Conference on Learning Representations, ICLR 2017\u2014Conference Track Proceedings, Toulon, France."}],"container-title":["Remote Sensing"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2072-4292\/14\/3\/644\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,10]],"date-time":"2025-10-10T22:10:34Z","timestamp":1760134234000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2072-4292\/14\/3\/644"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,1,29]]},"references-count":40,"journal-issue":{"issue":"3","published-online":{"date-parts":[[2022,2]]}},"alternative-id":["rs14030644"],"URL":"https:\/\/doi.org\/10.3390\/rs14030644","relation":{},"ISSN":["2072-4292"],"issn-type":[{"type":"electronic","value":"2072-4292"}],"subject":[],"published":{"date-parts":[[2022,1,29]]}}}