Towards advancing the earthquake forecasting by machine learning of satellite data

Xiong, Pan; Tong, Lei; Zhang, Kun; Shen, Xuhui; Battiston, R.; Ouzounov, Dimitar; Iuppa, R.; Crookes, Danny; Long, Cheng; Zhou, Huiyu

doi:10.1016/j.scitotenv.2021.145256

Cited by 52 publications

(23 citation statements)

References 63 publications

(46 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…dust storms [42], volcanic eruptions [43], etc. ), thereby the previous monitoring of tropospheric ozone is based on comparative analysis of long time-series data [44][45][46]. In particular, the research by Piscini et al shows that the total column ozone was less statistically significant through analyzing long time-series data [44], and our study indicates that the anomalies of tropospheric ozone in short time-series are also not evident.…”

Section: Discussioncontrasting

confidence: 42%

Atmospheric Anomaly Analysis Related to Ms > 6.0 Earthquakes in China during 2020–2021

Xu¹,

Chen²,

Yu³

et al. 2021

Remote Sensing

View full text Add to dashboard Cite

The attention towards links of atmospheric parameter variation and earthquakes has increased exponentially by utilizing new methods and more accurate observations. Persistent research makes it possible to gain insight into the precursor mechanism of earthquakes. In this paper, we studied the universality of detecting atmospheric anomalies associated with earthquakes based on tidal force fluctuation in China for earthquakes of Ms > 6.0, and explored the influence of tidal force on tectonic stress. The data of air temperature, geopotential height, ozone mixing ratio, and relative humidity from the National Center for Environmental Prediction (NCEP) were analyzed to reveal the spatiotemporal variation of atmospheric anomalies at multiple isobaric surfaces. Furthermore, the coupling of atmospheric parameters was investigated. The results showed that continuous solicitation exerted by tidal forces could change the strength of tectonic stress that causes earthquakes. The evolution pattern of air temperature, geopotential height, and relative humidity could be supported by atmospheric thermal vertical diffusion, while the anomalies of ozone mixing ratio was not evident. This verified the feasibility of detecting multi-parameter atmospheric anomalies associated with earthquakes based on tidal force fluctuation. Our results provide more evidence for understanding the atmospheric precursor characteristics of earthquakes.

show abstract

Section: Discussioncontrasting

confidence: 42%

Atmospheric Anomaly Analysis Related to Ms > 6.0 Earthquakes in China during 2020–2021

Xu¹,

Chen²,

Yu³

et al. 2021

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Deep learning, which has been widely used in recent earthquake research [18][19][20][21][22], could perform consistent analysis and assessment of a large number of earthquakes, and could take into account the influence of non-seismic anomalies, which are an effective tool to solve the above issue. By investigating different DEMETER satellite datasets, Xiong et al [21] confirmed some frequency bands with low-frequency electric and magnetic fields to be the main features for pre-seismic electromagnetic perturbation identification using deep learning.…”

Section: Introductionmentioning

confidence: 99%

“…Xiong et al [23] also proposed a deep learning framework termed SeqNetQuake by training whole life cycle dataset from the DEMETER satellites and transferring the well-trained model to the CSES satellite to form a new identification model which achieved a 12% improvement in classification performance. Based on the classical AdaBoost machine learning algorithm and the feature of satellite remote sensing products such as infrared and hyperspectral gases, Xiong et al [22] proposed a novel earthquake prediction framework based on inverse boosting pruning trees (IBPT), and achieved promising forecasting results in the validation of global earthquake cases retrospectively.…”

Section: Introductionmentioning

confidence: 99%

SafeNet: SwArm for Earthquake Perturbations Identification Using Deep Learning Networks

et al. 2021

Self Cite

View full text Add to dashboard Cite

Low Earth orbit satellites collect and study information on changes in the ionosphere, which contributes to the identification of earthquake precursors. Swarm, the European Space Agency three-satellite mission, has been launched to monitor the Earth geomagnetic field, and has successfully shown that in some cases it is able to observe many several ionospheric perturbations that occurred as a result of large earthquake activity. This paper proposes the SafeNet deep learning framework for detecting pre-earthquake ionospheric perturbations. We trained the proposed model using 9017 recent (2014–2020) independent earthquakes of magnitude 4.8 or greater, as well as the corresponding 7-year plasma and magnetic field data from the Swarm A satellite, and excellent performance has been achieved. In addition, the influence of different model inputs and spatial window sizes, earthquake magnitudes, and daytime or nighttime was explored. The results showed that for electromagnetic pre-earthquake data collected within a circular region of the epicenter and with a Dobrovolsky-defined radius and input window size of 70 consecutive data points, nighttime data provided the highest performance in discriminating pre-earthquake perturbations, yielding an F1 score of 0.846 and a Matthews correlation coefficient of 0.717. Moreover, SafeNet performed well in identifying pre-seismic ionospheric anomalies with increasing earthquake magnitude and unbalanced datasets. Hypotheses on the physical causes of earthquake-induced ionospheric perturbations are also provided. Our results suggest that the performance of pre-earthquake ionospheric perturbation identification can be significantly improved by utilizing SafeNet, which is capable of detecting precursor effects within electromagnetic satellite data.

show abstract

“…Recently some authors have explored the application of machine learning methods for automatically classifying and recognizing earthquake precursors on ground and in space (see for example Rouet-Leduc et al, 2017;Li et al, 2018;Akyol et al, 2020;Johnson et al, 2021;Xiong et al, 2020;Xiong et al, 2021). The field of research is very interesting, but asks for some caution.…”

Section: Rejecting Non-earthquake Associated Effects From Data Analysismentioning

confidence: 99%

A Critical Review of Ground Based Observations of Earthquake Precursors

2021

View full text Add to dashboard Cite

We aim at giving a short review of the seismo-associated phenomena detected on ground that in recent years have been investigated as possible earthquake precursors. The paper comes together with a companion article–published on this same volume by Picozza et al., 2021–devoted to summarize the space-based observation of earthquake–precursors by satellites missions. In the present work, we give an overview of the observations carried out on ground in order to identify earthquake precursors by distinguishing them from the large background constituted by both natural non-seismic and artificial sources. We start discussing the measurements of mechanical parameters and variations of geochemical fluids detected before earthquakes; then we review thermal and atmospheric oscillations; finally, observations of electromagnetic and ionospheric parameters possibly related to the occurrence of impeding earthquakes are discussed. In order to introduce a so large field of research, we focus only on some main case studies and statistical analyses together with the main hypotheses and models proposed in literature in order to explain the observed phenomenology.

show abstract

Towards advancing the earthquake forecasting by machine learning of satellite data

Cited by 52 publications

References 63 publications

Atmospheric Anomaly Analysis Related to Ms > 6.0 Earthquakes in China during 2020–2021

Atmospheric Anomaly Analysis Related to Ms > 6.0 Earthquakes in China during 2020–2021

SafeNet: SwArm for Earthquake Perturbations Identification Using Deep Learning Networks

A Critical Review of Ground Based Observations of Earthquake Precursors

Contact Info

Product

Resources

About