Double-Step U-Net: A Deep Learning-Based Approach for the Estimation of Wildfire Damage Severity through Sentinel-2 Satellite Data
Source abstract
Abstract
Wildfire damage severity census is a crucial activity for estimating monetary losses and for planning a prompt restoration of the affected areas. It consists in assigning, after a wildfire, a numerical damage/severity level, between 0 and 4, to each sub-area of the hit area. While burned area identification has been automatized by means of machine learning algorithms, the wildfire damage severity census operation is usually still performed manually and requires a significant effort of domain experts through the analysis of imagery and, sometimes, on-site missions. In this paper, we propose a novel supervised learning approach for the automatic estimation of the damage/severity level of the hit areas after the wildfire extinction. Specifically, the proposed approach, leveraging on the combination of a classification algorithm and a regression one, predicts the damage/severity level of the sub-areas of the area under analysis by processing a single post-fire satellite acquisition. Our approach has been validated in five different European countries and on 21 wildfires. It has proved to be robust for the application in several geographical contexts presenting similar geological aspects.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Fire effects on ecosystems · Atmospheric and Environmental Gas Dynamics · Fire Detection and Safety Systems
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Politecnico di Torino, corso Duca degli Abruzzi, 24, 10129 Torino, Italy
LINKS Foundation, via Pier Carlo Boggio, 61, 10138 Torino, Italy
Funding
Horizon 2020
References (56)
Show source-provided references
1. European Commission (2020, February 22). More Countries Than Ever Hit by Forest Fires in 2018. Available online: https://ec.europa.eu/jrc/en/news/more-countries-ever-hit-forest-fires-2018.
2. European Union (2020, February 20). Copernicus Sentinel-2 Mission. Available online: https://emergency.copernicus.eu/.
3. Near Real-Time Wildfire Progression Monitoring with Sentinel-1 SAR Time Series and Deep Learning10.1038/s41598-019-56967-x ↗
4. A deep learning approach for mapping and dating burned areas using temporal sequences of satellite images10.1016/j.isprsjprs.2019.12.014 ↗
5. The Collection 6 MODIS burned area mapping algorithm and product10.1016/j.rse.2018.08.005 ↗
6. Semi-automated mapping of burned areas in semi-arid ecosystems using MODIS time-series imagery
7. Satellite-based automated burned area detection: A performance assessment of the MODIS MCD45A1 in the Brazilian savanna
8. Hughes, M., Kaylor, S., and Hayes, D. (2017). Patch-based forest change detection from Landsat time series. Forests, 8.10.3390/f8050166 ↗
9. MODIS–Landsat fusion for large area 30 m burned area mapping10.1016/j.rse.2015.01.022 ↗
10. Ramo, R., and Chuvieco, E. (2017). Developing a random forest algorithm for MODIS global burned area classification. Remote Sens., 9.10.3390/rs9111193 ↗
11. A data mining approach for global burned area mapping
12. Shan, T., Wang, C., Chen, F., Wu, Q., Li, B., Yu, B., Shirazi, Z., Lin, Z., and Wu, W. (2017). A Burned Area Mapping Algorithm for Chinese FengYun-3 MERSI Satellite Data. Remote Sens., 9.10.3390/rs9070736 ↗
13. Bin, W., Ming, L., Dan, J., Suju, L., Qiang, C., Chao, W., Yang, Z., Huan, Y., and Jun, Z. (August, January 28). A Method of Automatically Extracting Forest Fire Burned Areas Using Gf-1 Remote Sensing Images. Proceedings of the IGARSS 2019-2019 IEEE International Geoscience and Remote Sensing Symposium, Yokohama, Japan.10.1109/IGARSS.2019.8900399 ↗
14. Verhegghen, A., Eva, H., Ceccherini, G., Achard, F., Gond, V., Gourlet-Fleury, S., and Cerutti, P. (2016). The potential of Sentinel satellites for burnt area mapping and monitoring in the Congo Basin forests. Remote Sens., 8.10.3390/rs8120986 ↗
15. Landsat-8 and Sentinel-2 burned area mapping-A combined sensor multi-temporal change detection approach10.1016/j.rse.2019.111254 ↗
16. Development of a Sentinel-2 burned area algorithm: Generation of a small fire database for sub-Saharan Africa10.1016/j.rse.2018.12.011 ↗
17. Stavrakoudis, D., Katagis, T., Minakou, C., and Gitas, I.Z. (2019, January 18–21). Towards a fully automatic processing chain for operationally mapping burned areas countrywide exploiting Sentinel-2 imagery. Proceedings of the Seventh International Conference on Remote Sensing and Geoinformation of the Environment (RSCy2019). International Society for Optics and Photonics, Paphos, Cyprus.10.1117/12.2535816 ↗
18. Filipponi, F. (2019). Exploitation of Sentinel-2 Time Series to Map Burned Areas at the National Level: A Case Study on the 2017 Italy Wildfires. Remote Sens., 11.10.3390/rs11060622 ↗
19. Farasin, A., Nini, G., Garza, P., and Rossi, C. (2020, June 23). Unsupervised Burned Area Estimation through Satellite Tiles: A Multimodal Approach by Means of Image Segmentation Over Remote Sensing Imagery. CEUR-WS. Available online: http://ceur-ws.org/Vol-2466/paper7.pdf.
20. Farasin, A., Colomba, L., Palomba, G., Nini, G., and Rossi, C. (2020, January 24–27). Supervised Burned Areas delineation by means of Sentinel-2 imagery and Convolutional Neural Networks. Proceedings of the 17th International Conference on Information Systems for Crisis Response and Management (ISCRAM 2020), Virginia Tech, Blacksburg, VA, USA.
21. European Union (2020, January 18). Copernicus EMS, Service Overview. Available online: https://emergency.copernicus.eu/mapping/ems/service-overview.
22. Quantifying burn severity in a heterogeneous landscape with a relative version of the delta Normalized Burn Ratio (dNBR)10.1016/j.rse.2006.12.006 ↗
23. Lutes, D.C., Keane, R.E., Caratti, J.F., Key, C.H., Benson, N.C., Steve, S., and Gangi, L.J. (2006). Landscape assessment (LA). FIREMON: Fire Effects Monitoring and Inventory System, Department of Agriculture, Forest Service, Rocky Mountain Research Station. Gen. Tech. Rep. RMRS-GTR-164-CD.10.2737/RMRS-GTR-164 ↗
24. Evaluation of forest fire on Madeira Island using Sentinel-2A MSI imagery
25. Saulino, L., Rita, A., Migliozzi, A., Maffei, C., Allevato, E., Garonna, A.P., and Saracino, A. (2020). Detecting Burn Severity across Mediterranean Forest Types by Coupling Medium-Spatial Resolution Satellite Imagery and Field Data. Remote Sens., 12.10.3390/rs12040741 ↗
26. Franco, M.G., Mundo, I.A., and Veblen, T.T. (2020). Field-Validated Burn-Severity Mapping in North Patagonian Forests. Remote Sens., 12.10.3390/rs12020214 ↗
27. Estimating burn severity and carbon emissions from a historic megafire in boreal forests of China10.1016/j.scitotenv.2020.136534 ↗
28. Zheng, Z., Wang, J., Shan, B., He, Y., Liao, C., Gao, Y., and Yang, S. (2020). A New Model for Transfer Learning-Based Mapping of Burn Severity. Remote Sens., 12.10.3390/rs12040708 ↗
29. A remote sensing approach to mapping fire severity in south-eastern Australia using sentinel 2 and random forest10.1016/j.rse.2020.111702 ↗
30. Copernicus, European Space Agency (ESA) (2020, April 13). Copernicus Sentinel-2 Mission. Available online: https://sentinel.esa.int/web/sentinel/missions/sentinel-2.
31. Copernicus, European Space Agency (ESA) (2020, April 12). Sentinel-2 Level-1C Products. Available online: https://earth.esa.int/web/sentinel/user-guides/sentinel-2-msi/product-types/level-1c.
32. Copernicus, European Space Agency (ESA) (2020, April 13). Sentinel-2 Level-1C Processing. Available online: https://earth.esa.int/web/sentinel/technical-guides/sentinel-2-msi/level-1c-processing.
33. Copernicus, European Space Agency (ESA) (2020, April 13). Sentinel-2 Level-2A Products. Available online: https://sentinel.esa.int/web/sentinel/user-guides/sentinel-2-msi/processing-levels/level-2.
34. Copernicus, European Space Agency (ESA) (2020, April 11). Sentinel-2 Level-2A Processing. Available online: https://earth.esa.int/web/sentinel/technical-guides/sentinel-2-msi/level-2a/algorithm.
35. Sinergise (2020, January 23). Sentinel-Hub platform. Available online: https://www.sentinel-hub.com/.
36. Automated cloud and cloud shadow identification in Landsat MSS imagery for temperate ecosystems10.1016/j.rse.2015.08.006 ↗
37. Detection of GNSS ionospheric scintillations based on machine learning decision tree10.1109/TAES.2018.2850385 ↗
38. Filipponi, F. (2018). BAIS2: Burned Area Index for Sentinel-2. Multidiscip. Digit. Publ. Inst. Proc., 2.10.3390/ecrs-2-05177 ↗
39. Remote sensing of fire severity: Assessing the performance of the normalized burn ratio10.1109/LGRS.2005.858485 ↗
40. Zanetti, M., Marinelli, D., Bertoluzza, M., Saha, S., Bovolo, F., Bruzzone, L., Magliozzi, M.L., Zavagli, M., and Costantini, M. (2019, January 5–7). A high resolution burned area detector for Sentinel-2 and Landsat-8. Proceedings of the 2019 10th International Workshop on the Analysis of Multitemporal Remote Sensing Images (MultiTemp), Shanghai, China.10.1109/Multi-Temp.2019.8866958 ↗
41. Evaluating the capabilities of Sentinel-2 for quantitative estimation of biophysical variables in vegetation10.1016/j.isprsjprs.2013.04.007 ↗
42. Yang, X., Zhao, S., Qin, X., Zhao, N., and Liang, L. (2017). Mapping of urban surface water bodies from Sentinel-2 MSI imagery at 10 m resolution via NDWI-based image sharpening. Remote Sens., 9.10.3390/rs9060596 ↗
43. Ronneberger, O., Fischer, P., and Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. International Conference on Medical Image Computing And Computer-Assisted Intervention, Springer.10.1007/978-3-319-24574-4_28 ↗
44. Long, J., Shelhamer, E., and Darrell, T. (2015, January 7–12). Fully convolutional networks for semantic segmentation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Boston, MA, USA.10.1109/CVPR.2015.7298965 ↗
45. Automated deep learning method for whole-breast segmentation in diffusion-weighted breast MRI10.1002/jmri.26860 ↗
46. Multilevel UNet for pancreas segmentation from non-contrast CT scans through domain adaptation
47. Dutta, J., Chakraborty, D., and Mondal, D. (2020). Multimodal Segmentation of Brain Tumours in Volumetric MRI Scans of the Brain Using Time-Distributed U-Net. Computational Intelligence in Pattern Recognition, Springer.10.1007/978-981-13-9042-5_62 ↗
48. Goodfellow, I., Bengio, Y., and Courville, A. (2016). Deep Learning, MIT Press.
49. GDAL/OGR CContributors (2020). GDAL/OGR Geospatial Data Abstraction Software Library, Open Source Geospatial Foundation.10.22224/gistbok/2020.4.1 ↗
50. Scikit-image: Image processing in Python10.7717/peerj.453 ↗
51. Buitinck, L., Louppe, G., Blondel, M., Pedregosa, F., Mueller, A., Grisel, O., Niculae, V., Prettenhofer, P., Gramfort, A., and Grobler, J. (2013, January 23–27). API design for machine learning software: Experiences from the scikit-learn project. Proceedings of the ECML PKDD Workshop: Languages for Data Mining and Machine Learning, Prague, Czech Republic.
52. Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., Killeen, T., Lin, Z., Gimelshein, N., and Antiga, L. (2019). PyTorch: An Imperative Style, High-Performance Deep Learning Library. Advances in Neural Information Processing Systems 32, Curran Associates, Inc.
53. Ng, A. (2004). Proceedings of the Twenty-First International Conference on Machine Learning, Association for Computing Machinery.
54. Soomro, T.A., Hellwich, O., Afifi, A.J., Paul, M., Gao, J., and Zheng, L. (2018, January 10–13). Strided U-Net model: Retinal vessels segmentation using dice loss. Proceedings of the 2018 IEEE Digital Image Computing: Techniques and Applications (DICTA), Canberra, Australia.10.1109/DICTA.2018.8615770 ↗
55. Glorot, X., and Bengio, Y. (2010, January 13–15). Understanding the difficulty of training deep feedforward neural networks. Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, Sardinia, Italy.
56. Flach, P. (2012). Machine Learning: The Art and Science of Algorithms That Make Sense of Data, Cambridge University Press.10.1017/CBO9780511973000 ↗