Priyadarshana, Anuradha, and Das, Sourav (2021) Machine based algorithm for characterizing and precursory monitoring of landslides. In: [Presented at the 7th International Conference on Time Series and Forecasting]. From: ITISE 2021: 7th International Conference on Time Series and Forecasting, 19-21 July 2021, Gran Canaria, Spain and Online.

Preview

PDF (Abstract Only) - Accepted Version Download (76kB) | Preview

Abstract

Landslides routinely cause significant damage to life and property. Various studies have evaluated mechanisms that trigger landslides. It is now established that a large proportion landslides are caused by a combination of anthropogenic activities and environmental factors such as rain and earthquakes. Das and Tordesillas (2019) proposed a machine based algorithm to characterize and provide precursory warnings for a rockfall type of landslide. In this work we extend this algorithm (Section 3.4) to include relatively short but serially correlated displacement signals sampled with using the interferometric synthetic aperture radars (InSAR) embeded within the sentinel group of satellites. The generalized version of the algorithm is based on spectral analysis of time series. The resulting algorithm is applied to 2019 Brumadinho Tailing Dam collapse. The algorithm identities several potential risk milestones going back to about July, 2018, eventually concluding with two retrospective estimates of risk, tR (27 Feb 2018 - 26 Aug 2018) for definitive but emergent risk and tI (27 Jun 2018 - 24 Dec 2018) for imminent risk of collapse of the system. We posit that the combination of spectral methods and second order statistical properties of displacement signals can reveal tantalizing signs of transition into an unstable regime and argue that this algorithm can also be used for intervention with a view to mitigate the chances of a potential disaster.

Item ID:	70816
Item Type:	Conference Item (Abstract / Summary)
Related URLs:	https://itise.ugr.es/ProgramITISE2021.pd...
Date Deposited:	15 Dec 2021 22:38
FoR Codes:	37 EARTH SCIENCES > 3704 Geoinformatics > 370499 Geoinformatics not elsewhere classified @ 30% 49 MATHEMATICAL SCIENCES > 4905 Statistics > 490511 Time series and spatial modelling @ 40% 40 ENGINEERING > 4013 Geomatic engineering > 401302 Geospatial information systems and geospatial data modelling @ 30%
SEO Codes:	19 ENVIRONMENTAL POLICY, CLIMATE CHANGE AND NATURAL HAZARDS > 1904 Natural hazards > 190403 Geological hazards (e.g. earthquakes, landslides and volcanic activity) @ 100%
Downloads:	Total: 44 Last 12 Months: 3
	More Statistics