Machine learning and best fit approach to map lava flows from space

Estimating the areal coverage of newly erupted lava is both a crucial component of volcano monitoring and a powerful tool for characterizing lava flow emplacement behavior. Here, it is presented a methodology based on machine learning, developed in the Google Earth Engine platform, and best fitting approach, to map the extent of lava flows solely using freely available and open-source data from space-borne instruments. Radar and optical satellite data are used as input to different machine learning techniques and best fitting models, so that the methodology is able to operate in all weather conditions. The satellite-driven approach has been successfully used for mapping lava flows automatically during the long sequence of summit eruptions occurred at Mt. Etna between December 2020 and October 2021.

[1] Del Negro C. et al., Sci. Rep., 3 (2013) 1.
[2] Del Negro C. et al., Geol. Soc. Am. Bull., 132 (2020) 1615.
[3] Harris A., Thermal Remote Sensing of Active Volcanoes (Cambridge University Press) 2013, https://doi.org/10.1017/CBO9781139029346.
[4] Corradino C. et al., Remote Sens., 11 (2019) 1916.
[5] Corradino C. et al., Energies, 14 (2021) 197.
[6] Corradino C. et al., Remote Sens., 13 (2021) 4080.
[7] Amato E. et al., in 2021 International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME) (IEEE) 2021, p. 1, https:// doi.org/10.1109/ICECCME52200.2021.9591110.
[8] Corradino C. et al., in 2021 17th International Workshop on Cellular Nanoscale Networks and their Applications (CNNA) (IEEE) 2021, p. 1, https://doi.org/10.1109/CNNA49188.2021.9610813.
[9] Amato E. et al., Earth and Space Science Open Archive (2021), https://doi.org/10.1002/essoar.10509929.1.
[10] Torrisi F. et al., Earth and Space Science Open Archive (2021), https://doi.org/10.1002/essoar.10509947.1.
[11] Corradino C. et al., Earth and Space Science Open Archive (2021), https://doi.org/10.1002/essoar.10510119.1.
[12] Lu Z. et al., Remote Sens. Environ., 91 (2004) 345.
[13] Freeman E. A. et al., Ecol. Model., 217 (2008) 48.
[14] Bonaccorso G., Machine Learning Algorithms (Packt Publishing Ltd.) 2017.
[15] Rogic N. et al., Remote Sens., 11 (2019) 662.
[16] Corradino C. et al., Remote Sens., 12 (2020) 970.
[17] Ganci G. et al., Geophys. Res. Lett., 39 (2012) L06305.
[18] Gorelick N. et al., Remote Sens. Environ., 202 (2017) 18.
[19] Vicari A. et al., Ann. Geophys., (2011) https://doi.org/10.4401/ag-5347.
[20] Zerrouki N. et al., in 2014 IEEE International Conference on Systems, Man, and Cybernetics (SMC) (IEEE) 2014, p. 864, https://doi.org/10.1109/SMC.2014.6974020.
[21] Géron A., Hands-on Machine Learning with Scikit-Learn, Keras, and TensorFlow: Concepts, Tools, and Techniques to Build Intelligent Systems (O’Reilly Media, Inc.) 2019.
[22] Marchese F. et al., Remote Sens., 13 (2021) 3074.
[23] Corradino C. et al., Energies, 12 (2019) 1216.
[24] Wooster M. J. et al., Remote Sens. Environ., 86 (2003) 83.
[25] Wooster M. J. et al., J. Geophys. Res.: Atmos., 110 (2005) D24311.
[26] Zhang T. et al., Remote Sens. Environ., 198 (2017) 407.
[27] Xu W. et al., Remote Sens. Environ., 261 (2021) 112460.
[28] Wadge G. et al., J. Volcanol. Geotherm. Res., 199 (2011) 142.
[29] Wadge G. et al., Geol. Soc. London Mem., 21 (2002) 583.
[30] Arnold D. et al., Remote Sens. Environ., 209 (2018) 480.