{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,23]],"date-time":"2026-02-23T21:48:13Z","timestamp":1771883293677,"version":"3.50.1"},"reference-count":56,"publisher":"MDPI AG","issue":"8","license":[{"start":{"date-parts":[[2023,4,18]],"date-time":"2023-04-18T00:00:00Z","timestamp":1681776000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"Australian Government Research Training Program (RTP) Scholarship"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Remote Sensing"],"abstract":"<jats:p>Satellite imagery is the only feasible approach to annual monitoring and reporting on land cover change. Unfortunately, conventional pixel-based classification methods based on spectral response only (e.g., using random forests algorithms) have shown a lack of spatial and temporal stability due, for instance, to variability between individual pixels and changes in vegetation condition, respectively. Machine learning methods that consider spatial patterns in addition to reflectance can address some of these issues. In this study, a convolutional neural network (CNN) model, U-Net, was trained for a 500 km \u00d7 500 km region in southeast Australia using annual Landsat geomedian data for the relatively dry and wet years of 2018 and 2020, respectively. The label data for model training was an eight-class classification inferred from a static land-use map, enhanced using forest-extent mapping. Here, we wished to analyse the benefits of CNN-based land cover mapping and reporting over 34 years (1987\u20132020). We used the trained model to generate annual land cover maps for a 100 km \u00d7 100 km tile near the Australian Capital Territory. We developed innovative diagnostic methods to assess spatial and temporal stability, analysed how the CNN method differs from pixel-based mapping and compared it with two reference land cover products available for some years. Our U-Net CNN results showed better spatial and temporal stability with, respectively, overall accuracy of 89% verses 82% for reference pixel-based mapping, and 76% of pixels unchanged over 33 years. This gave a clearer insight into where and when land cover change occurred compared to reference mapping, where only 30% of pixels were conserved. Remaining issues include edge effects associated with the CNN method and a limited ability to distinguish some land cover types (e.g., broadacre crops vs. pasture). We conclude that the CNN model was better for understanding broad-scale land cover change, use in environmental accounting and natural resource management, whereas pixel-based approaches sometimes more accurately represented small-scale changes in land cover.<\/jats:p>","DOI":"10.3390\/rs15082132","type":"journal-article","created":{"date-parts":[[2023,4,19]],"date-time":"2023-04-19T01:09:22Z","timestamp":1681866562000},"page":"2132","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":10,"title":["Convolutional Neural Network Shows Greater Spatial and Temporal Stability in Multi-Annual Land Cover Mapping Than Pixel-Based Methods"],"prefix":"10.3390","volume":"15","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-1797-2255","authenticated-orcid":false,"given":"Tony","family":"Boston","sequence":"first","affiliation":[{"name":"Fenner School of Environment and Society, Australian National University, Acton, ACT 2601, Australia"}]},{"given":"Albert","family":"Van Dijk","sequence":"additional","affiliation":[{"name":"Fenner School of Environment and Society, Australian National University, Acton, ACT 2601, Australia"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9607-3751","authenticated-orcid":false,"given":"Richard","family":"Thackway","sequence":"additional","affiliation":[{"name":"Fenner School of Environment and Society, Australian National University, Acton, ACT 2601, Australia"}]}],"member":"1968","published-online":{"date-parts":[[2023,4,18]]},"reference":[{"key":"ref_1","unstructured":"Cresswell, I.D., Janke, T., and Johnston, E.L. (2021). Australia State of the Environment 2021: Overview, Commonwealth of Australia. Independent Report to the Australian Government Minister for the Environment."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"e117","DOI":"10.1111\/csp2.117","article-title":"Lots of Loss with Little Scrutiny: The Attrition of Habitat Critical for Threatened Species in Australia","volume":"1","author":"Ward","year":"2019","journal-title":"Conserv. Sci Pract."},{"key":"ref_3","doi-asserted-by":"crossref","unstructured":"Thackway, R. (2018). Land Use in Australia: Past, Present and Future, ANU Press. Available online: https:\/\/press.anu.edu.au\/publications\/land-use-australia.","DOI":"10.22459\/LUA.02.2018"},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"367","DOI":"10.1046\/j.1365-2699.1999.00139.x","article-title":"Global Land Cover Characterization from Satellite Data: From Research to Operational Implementation? GCTE\/LUCC Research Review","volume":"8","author":"Defries","year":"1999","journal-title":"Glob. Ecol. Biogeogr."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"5","DOI":"10.1023\/A:1010933404324","article-title":"Random Forests","volume":"45","author":"Breiman","year":"2001","journal-title":"Mach. Learn."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"725","DOI":"10.1080\/01431160110040323","article-title":"An Assessment of Support Vector Machines for Land Cover Classification","volume":"23","author":"Huang","year":"2002","journal-title":"Int. J. Remote Sens."},{"key":"ref_7","unstructured":"Lymburner, L., Tan, P., Mueller, N., Thackway, R., Lewis, A., Thankappan, M., Randall, L., Islam, A., and Senarath, U. (2011). The National Dynamic Land Cover Dataset 2011, Geoscience Australia, ACT."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"1715","DOI":"10.1080\/17538947.2022.2130461","article-title":"Operational Continental-Scale Land Cover Mapping of Australia Using the Open Data Cube","volume":"15","author":"Owers","year":"2022","journal-title":"Int. J. Digit. Earth"},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"112148","DOI":"10.1016\/j.rse.2020.112148","article-title":"High-Resolution Wall-to-Wall Land-Cover Mapping and Land Change Assessment for Australia from 1985 to 2015","volume":"252","author":"Hadjikakou","year":"2021","journal-title":"Remote Sens. Environ."},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"3277","DOI":"10.1080\/01431161.2020.1871094","article-title":"Fast and Accurate Land-Cover Classification on Medium-Resolution Remote-Sensing Images Using Segmentation Models","volume":"42","author":"Zhang","year":"2021","journal-title":"Int. J. Remote Sens."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"306","DOI":"10.1007\/978-3-030-99170-8_22","article-title":"Land Cover Classification Using CNN and Semantic Segmentation: A Case of Study in Antioquia, Colombia","volume":"Volume 1532","author":"Morillo","year":"2022","journal-title":"Smart Technologies, Systems and Applications"},{"key":"ref_12","unstructured":"Ulmas, P., and Liiv, I. (2020). Segmentation of Satellite Imagery using U-Net Models for Land Cover Classification. arXiv."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"166","DOI":"10.1016\/j.isprsjprs.2019.04.015","article-title":"Deep Learning in Remote Sensing Applications: A Meta-Analysis and Review","volume":"152","author":"Ma","year":"2019","journal-title":"ISPRS J. Photogramm. Remote Sens."},{"key":"ref_14","doi-asserted-by":"crossref","unstructured":"Boston, T., Van Dijk, A., Larraondo, P.R., and Thackway, R. (2022). Comparing CNNs and Random Forests for Landsat Image Segmentation Trained on a Large Proxy Land Cover Dataset. Remote Sens., 14.","DOI":"10.3390\/rs14143396"},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"8","DOI":"10.1109\/MGRS.2017.2762307","article-title":"Deep Learning in Remote Sensing: A Comprehensive Review and List of Resources","volume":"5","author":"Zhu","year":"2017","journal-title":"IEEE Geosci. Remote Sens. Mag."},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Hoeser, T., and Kuenzer, C. (2020). Object Detection and Image Segmentation with Deep Learning on Earth Observation Data: A Review-Part I: Evolution and Recent Trends. Remote Sens., 12.","DOI":"10.3390\/rs12101667"},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"24","DOI":"10.1016\/j.isprsjprs.2020.12.010","article-title":"Review on Convolutional Neural Networks (CNN) in vegetation remote sensing","volume":"173","author":"Kattenborn","year":"2021","journal-title":"ISPRS J. Photogramm. Remote Sens."},{"key":"ref_18","first-page":"555","article-title":"Object-Oriented Image Processing in an Integrated GIS\/Remote Sensing Environment and Perspectives for Environmental Applications","volume":"2","author":"Blaschke","year":"2000","journal-title":"Environ. Inf. Plan. Politics Public"},{"key":"ref_19","doi-asserted-by":"crossref","unstructured":"Phiri, D., and Morgenroth, J. (2017). Developments in Landsat Land Cover Classification Methods: A Review. Remote Sens., 9.","DOI":"10.3390\/rs9090967"},{"key":"ref_20","doi-asserted-by":"crossref","unstructured":"Zhang, X., Han, L., Han, L., and Zhu, L. (2020). How Well Do Deep Learning-Based Methods for Land Cover Classification and Object Detection Perform on High Resolution Remote Sensing Imagery?. Remote Sens., 12.","DOI":"10.3390\/rs12030417"},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"55","DOI":"10.1016\/j.isprsjprs.2016.03.008","article-title":"Optical Remotely Sensed Time Series Data for Land Cover Classification: A Review","volume":"116","author":"White","year":"2016","journal-title":"ISPRS J. Photogramm. Remote Sens."},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"156","DOI":"10.1016\/j.rse.2016.10.010","article-title":"Assessing the Robustness of Random Forests to Map Land Cover with High Resolution Satellite Image Time Series over Large Areas","volume":"187","author":"Pelletier","year":"2016","journal-title":"Remote Sens. Environ."},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"Stoian, A., Poulain, V., Inglada, J., Poughon, V., and Derksen, D. (2019). Land Cover Maps Production with High Resolution Satellite Image Time Series and Convolutional Neural Networks: Adaptations and Limits for Operational Systems. Remote Sens., 11.","DOI":"10.20944\/preprints201906.0270.v1"},{"key":"ref_24","first-page":"103","article-title":"Detection of Spatio-Temporal Evolutions on Multi-Annual Satellite Image Time Series: A Clustering Based Approach","volume":"74","author":"Khiali","year":"2019","journal-title":"Int. J. Appl. Earth Obs. Geoinf."},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"1331","DOI":"10.1080\/014311600210209","article-title":"Global Land Cover Classification at 1 Km Spatial Resolution Using a Classification Tree Approach","volume":"21","author":"Hansen","year":"2000","journal-title":"Int. J. Remote Sens."},{"key":"ref_26","doi-asserted-by":"crossref","first-page":"373","DOI":"10.1080\/17538947.2012.713190","article-title":"Global Characterization and Monitoring of Forest Cover Using Landsat Data: Opportunities and Challenges","volume":"5","author":"Townshend","year":"2012","journal-title":"Int. J. Digit. Earth"},{"key":"ref_27","unstructured":"Geoscience Australia (2022, November 01). Digital Earth Australia-Public Data-Land Cover (Landsat) v1.0.0, Available online: https:\/\/cmi.ga.gov.au\/data-products\/dea\/607\/dea-land-cover-landsat."},{"key":"ref_28","doi-asserted-by":"crossref","unstructured":"Karra, K., Kontgis, C., Statman-Weil, Z., Mazzariello, J.C., Mathis, M., and Brumby, S.P. (2021, January 11). Global Land Use\/Land Cover with Sentinel 2 and Deep Learning. Proceedings of the 2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS, Brussels, Belgium.","DOI":"10.1109\/IGARSS47720.2021.9553499"},{"key":"ref_29","unstructured":"Geoscience Australia (2022, November 01). Digital Earth Australia-Public Data-Surface Reflectance 25m Geomedian v2.1.0, Available online: https:\/\/data.dea.ga.gov.au\/?prefix=geomedian-australia\/v2.1.0\/."},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"6254","DOI":"10.1109\/TGRS.2017.2723896","article-title":"High-Dimensional Pixel Composites From Earth Observation Time Series","volume":"55","author":"Roberts","year":"2017","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"192","DOI":"10.1080\/07038992.2014.945827","article-title":"Pixel-Based Image Compositing for Large-Area Dense Time Series Applications and Science","volume":"40","author":"White","year":"2014","journal-title":"Can. J. Remote Sens."},{"key":"ref_32","doi-asserted-by":"crossref","unstructured":"Navab, N., Hornegger, J., Wells, W., and Frangi, A. (2015). Medical Image Computing and Computer-Assisted Intervention\u2014MICCAI 2015. MICCAI 2015, Springer. Lecture Notes in Computer Science.","DOI":"10.1007\/978-3-319-24553-9"},{"key":"ref_33","unstructured":"ABARES (2022, November 01). Catchment Scale Land Use of Australia\u2014Update December 2018, Available online: https:\/\/www.agriculture.gov.au\/abares\/aclump\/land-use\/catchment-scale-land-use-of-australia-update-december-2018."},{"key":"ref_34","unstructured":"ABARES (2022, November 01). Catchment Scale Land Use of Australia\u2014Update December 2020, Available online: https:\/\/www.agriculture.gov.au\/abares\/aclump\/catchment-scale-land-use-of-australia-update-december-2020."},{"key":"ref_35","unstructured":"ABARES (2022, November 01). Forests of Australia (2018), Available online: https:\/\/www.agriculture.gov.au\/abares\/forestsaustralia\/forest-data-maps-and-tools\/spatial-data\/forest-cover."},{"key":"ref_36","unstructured":"Kingma, D.P., and Ba, J. (2015, January 7\u20139). Adam: A Method for Stochastic Optimization. Proceedings of the 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA."},{"key":"ref_37","doi-asserted-by":"crossref","first-page":"576","DOI":"10.1002\/mp.13300","article-title":"AnatomyNet: Deep learning for fast and fully automated whole-volume segmentation of head and neck anatomy","volume":"46","author":"Zhu","year":"2019","journal-title":"Med. Phys."},{"key":"ref_38","doi-asserted-by":"crossref","unstructured":"Buslaev, A., Iglovikov, V.I., Khvedchenya, E., Parinov, A., Druzhinin, M., and Kalinin, A.A. (2020). Albumentations: Fast and Flexible Image Augmentations. Information, 11.","DOI":"10.3390\/info11020125"},{"key":"ref_39","doi-asserted-by":"crossref","first-page":"317","DOI":"10.1016\/j.rse.2005.08.006","article-title":"Land Cover Classification and Change Analysis of the Twin Cities (Minnesota) Metropolitan Area by Multitemporal Landsat Remote Sensing","volume":"98","author":"Yuan","year":"2005","journal-title":"Remote Sens. Environ."},{"key":"ref_40","doi-asserted-by":"crossref","unstructured":"Lucas, R., Mueller, N., Siggins, A., Owers, C., Clewley, D., Bunting, P., Kooymans, C., Tissott, B., Lewis, B., and Lymburner, L. (2019). Land Cover Mapping Using Digital Earth Australia. Data, 4.","DOI":"10.3390\/data4040143"},{"key":"ref_41","unstructured":"Di Gregorio, A., and Jansen, L.J.M. (2005). Land Cover Classification System: Classification Concepts and User Manual: LCCS, Software version 2, Food and Agriculture Organization of the United Nations."},{"key":"ref_42","doi-asserted-by":"crossref","first-page":"283","DOI":"10.1016\/S0001-2998(78)80014-2","article-title":"Basic Principles of ROC Analysis","volume":"8","author":"Metz","year":"1978","journal-title":"Semin. Nucl. Med."},{"key":"ref_43","doi-asserted-by":"crossref","first-page":"37","DOI":"10.1177\/001316446002000104","article-title":"A Coefficient of Agreement for Nominal Scales","volume":"20","author":"Cohen","year":"1960","journal-title":"Educ. Psychol. Meas."},{"key":"ref_44","doi-asserted-by":"crossref","first-page":"159","DOI":"10.2307\/2529310","article-title":"The Measurement of Observer Agreement for Categorical Data","volume":"33","author":"Landis","year":"1977","journal-title":"Biometrics"},{"key":"ref_45","unstructured":"ACT Government (2022, November 01). Canberra\u2019s Tree Canopy Continues to Grow, Available online: https:\/\/www.cmtedd.act.gov.au\/open_government\/inform\/act_government_media_releases\/chris-steel-mla-media-releases\/2021\/canberras-tree-canopy-continues-to-grow."},{"key":"ref_46","doi-asserted-by":"crossref","first-page":"15","DOI":"10.1080\/01431168708948612","article-title":"The Interactive Effect of Spatial Resolution and Degree of Internal Variability within Land-Cover Types on Classification Accuracies","volume":"8","author":"Cushnie","year":"1987","journal-title":"Int. J. Remote Sens."},{"key":"ref_47","doi-asserted-by":"crossref","first-page":"206","DOI":"10.1016\/S0034-4257(98)00112-6","article-title":"Fine Spatial Resolution Simulated Satellite Sensor Imagery for Land Cover Mapping in the United Kingdom","volume":"68","author":"Aplin","year":"1999","journal-title":"Remote Sens. Environ."},{"key":"ref_48","doi-asserted-by":"crossref","first-page":"2","DOI":"10.1016\/j.isprsjprs.2009.06.004","article-title":"Object Based Image Analysis for Remote Sensing","volume":"65","author":"Blaschke","year":"2010","journal-title":"ISPRS J. Photogramm. Remote Sens."},{"key":"ref_49","doi-asserted-by":"crossref","first-page":"259","DOI":"10.1016\/j.rse.2011.11.020","article-title":"A Comparison of Pixel-Based and Object-Based Image Analysis with Selected Machine Learning Algorithms for the Classification of Agricultural Landscapes Using SPOT-5 HRG Imagery","volume":"118","author":"Duro","year":"2012","journal-title":"Remote Sens. Environ."},{"key":"ref_50","doi-asserted-by":"crossref","unstructured":"Pelletier, C., Webb, G., and Petitjean, F. (2019). Temporal Convolutional Neural Network for the Classification of Satellite Image Time Series. Remote Sens., 11.","DOI":"10.3390\/rs11050523"},{"key":"ref_51","unstructured":"Guyon, I., Luxburg, U.V., Bengio, S., Wallach, H., Fergus, R., Vishwanathan, S., and Garnett, R. (2017). Proceedings of the Advances in Neural Information Processing Systems (NIPS 2017), Curran Associates, Inc.. Available online: https:\/\/proceedings.neurips.cc\/paper\/2017\/file\/3f5ee243547dee91fbd053c1c4a845aa-Paper.pdf."},{"key":"ref_52","unstructured":"Garnot, V.S.F., and Landrieu, L. (2021, January 11\u201317). Panoptic Segmentation of Satellite Image Time Series with Convolutional Temporal Attention Networks. Proceedings of the 2021 IEEE\/CVF International Conference on Computer Vision (ICCV), Montreal, QC, Canada."},{"key":"ref_53","doi-asserted-by":"crossref","unstructured":"Yao, J., and Jin, S. (2022). Multi-Category Segmentation of Sentinel-2 Images Based on the Swin UNet Method. Remote Sens., 14.","DOI":"10.3390\/rs14143382"},{"key":"ref_54","unstructured":"United Nations et al (2022, November 01). System of Environmental-Economic Accounting\u2014Ecosystem Accounting (SEEA EA). Available online: https:\/\/seea.un.org\/ecosystem-accounting."},{"key":"ref_55","doi-asserted-by":"crossref","first-page":"101413","DOI":"10.1016\/j.ecoser.2022.101413","article-title":"Establishing the SEEA Ecosystem Accounting as a Global Standard","volume":"54","author":"Edens","year":"2022","journal-title":"Ecosyst. Serv."},{"key":"ref_56","unstructured":"Australian Bureau of Statistics (2022, November 01). National Land Cover Account, ABS, Available online: https:\/\/www.abs.gov.au\/statistics\/environment\/environmental-management\/national-land-cover-account\/latest-release."}],"container-title":["Remote Sensing"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2072-4292\/15\/8\/2132\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,10]],"date-time":"2025-10-10T19:18:21Z","timestamp":1760123901000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2072-4292\/15\/8\/2132"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,4,18]]},"references-count":56,"journal-issue":{"issue":"8","published-online":{"date-parts":[[2023,4]]}},"alternative-id":["rs15082132"],"URL":"https:\/\/doi.org\/10.3390\/rs15082132","relation":{},"ISSN":["2072-4292"],"issn-type":[{"value":"2072-4292","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,4,18]]}}}