Terrain Proxy-Based Site Classification for Seismic Zonation in North Korea within a Geospatial Data-Driven Workflow

Numerous seismic activities occur in North Korea. However, it is difficult to perform seismic hazard assessment and obtain zonal data in the Korean Peninsula, including North Korea, when applying parametric or nonparametric methods. Remote sensing can be implemented for soil characterization or spatial zonation studies on irregular, surficial, and subsurface systems of inaccessible areas. Herein, a data-driven workflow for extracting the principal features using a digital terrain model (DTM) is proposed. In addition, geospatial grid information containing terrain features and the average shear wave velocity in the top 30 m of the subsurface (VS30) are employed using geostatistical interpolation methods; machine learning (ML)-based regression models were optimized and VS30-based seismic zonation in the test areas in North Korea were forecasted. The interrelationships between VS30 and terrain proxy (elevation, slope, and landform class) in the training area in South Korea were verified to define the input layer in regression models. The landform class represents a new proxy of VS30 and was subgrouped according to the correlation with grid-based VS30. The geospatial grid information was generated via the optimum geostatistical interpolation method (i.e., sequential Gaussian simulation (SGS)). The best-fitting model among four ML methods was determined by evaluating cost function-based prediction performance, performing uncertainty analysis for the empirical correlations of VS30, and studying spatial correspondence with the borehole-based VS30 map. Subsequently, the best-fitting regression models were designed by training the geospatial grid in South Korea. Then, DTM and its terrain features were constructed along with VS30 maps for three major cities (Pyongyang, Kaesong, and Nampo) in North Korea. A similar distribution of the VS30 grid obtained using SGS was shown in the multilayer perceptron-based VS30 map.

[1]  V. Vapnik.  The Support Vector Method of Function Estimation , 1998 .

[2]  Choong-Ki Chung,et al.  A Three-Dimensional Geotechnical Spatial Modeling Method for Borehole Dataset Using Optimization of Geostatistical Approaches , 2020 .

[3]  Teresa Gonçalves,et al.  Sentinel-2 Image Scene Classification: A Comparison between Sen2Cor and a Machine Learning Approach , 2021, Remote. Sens..

[4]  D. Wald,et al.  On the Use of High-Resolution Topographic Data as a Proxy for Seismic Site Conditions (VS30) , 2009 .

[5]  Jonathan P. Stewart,et al.  Compilation of a Local VS Profile Database and Its Application for Inference of VS30 from Geologic‐ and Terrain‐Based Proxies , 2014 .

[6]  Ünal Dikmen,et al.  Statistical correlations of shear wave velocity and penetration resistance for soils , 2009 .

[7]  조진철,et al.  Maritime Trade and Port Evolution in a Socialist Developing Country: Nampo, Gateway of North Korea , 2006 .

[8]  Chang-Guk Sun,et al.  GIS-based regional assessment of seismic site effects considering the spatial uncertainty of site-specific geotechnical characteristics in coastal and inland urban areas , 2017 .

[9]  Fan-Rui Meng,et al.  Predict soil texture distributions using an artificial neural network model , 2009 .

[10]  A. Seif.  Using Topography Position Index for Landform Classification (Case study: Grain Mountain) , 2015 .

[11]  F. Wenzel,et al.  Global Megathrust Earthquake Hazard—Maximum Magnitude Assessment Using Multi-Variate Machine Learning , 2019, Front. Earth Sci..

[12]  R. J. Pike,et al.  Automated classifications of topography from DEMs by an unsupervised nested-means algorithm and a three-part geometric signature , 2007 .

[13]  Shih-Wei Lin,et al.  Particle swarm optimization for parameter determination and feature selection of support vector machines , 2008, Expert Syst. Appl..

[14]  A. Skentos,et al.  Landform Analysis Using Terrain Attributes. A Gis Application on the Island of Ikaria (Aegean Sea, Greece) , 2017 .

[15]  Silvia Castellaro,et al.  Vs30: Proxy for Seismic Amplification? , 2008 .

[16]  Y. Ohta,et al.  Empirical shear wave velocity equations in terms of characteristic soil indexes , 1978 .

[17]  Jim Mori,et al.  Evaluation of Methods for Estimating Linear Site-Response Amplifications in the Los Angeles Region , 2000 .

[18]  Amir Mosavi,et al.  Bio-Inspired Hybridization of Artificial Neural Networks: An Application for Mapping the Spatial Distribution of Soil Texture Fractions , 2021, Remote. Sens..

[19]  G. Chae,et al.  Hydrochemistry of urban groundwater in Seoul, South Korea: effects of land-use and pollutant recharge , 2005 .

[20]  Jamie Doucette,et al.  Experimental territoriality: Assembling the Kaesong Industrial Complex in North Korea , 2015 .

[21]  Ryoji Iwasaki,et al.  ON DYNAMIC SHEAR MODULI AND POISSON'S RATIOS OF SOIL DEPOSITS , 1973 .

[22]  Chang-Guk Sun,et al.  Geospatial Assessment of the Post-Earthquake Hazard of the 2017 Pohang Earthquake Considering Seismic Site Effects , 2018, ISPRS Int. J. Geo Inf..

[23]  Maria Rosaria Gallipoli,et al.  Analysis and Modeling of hvsr in the Presence of a Velocity Inversion: The Case of Venosa, Italy , 2005 .

[24]  Jian Wang,et al.  Deep learning and its application in geochemical mapping , 2019, Earth-Science Reviews.

[25]  Kurt Hornik,et al.  Multilayer feedforward networks are universal approximators , 1989, Neural Networks.

[26]  Lei Zhao,et al.  The geology of North Korea: An overview , 2019, Earth-Science Reviews.

[27]  Seongryong Kim,et al.  1-D velocity model for the North Korean Peninsula from Rayleigh wave dispersion of ambient noise cross-correlations , 2019, Journal of Seismology.

[28]  Soo-Mi Choi,et al.  Ubiquitous GIS-Based Forest Fire Susceptibility Mapping Using Artificial Intelligence Methods , 2020, Remote. Sens..

[29]  S. Anitha,et al.  Application of a radial basis function neural network for diagnosis of diabetes mellitus , 2006 .

[30]  Jasper Snoek,et al.  Practical Bayesian Optimization of Machine Learning Algorithms , 2012, NIPS.

[31]  Paolo Tarolli,et al.  LiDAR-derived slopes for headwater channel network analysis , 2009 .

[32]  Walter L. Ruzzo,et al.  A Regression-based K nearest neighbor algorithm for gene function prediction from heterogeneous data , 2006, BMC Bioinformatics.

[33]  Bakhtiar Feizizadeh,et al.  DEM-Based Vs30 Map and Terrain Surface Classification in Nationwide Scale - A Case Study in Iran , 2019, ISPRS Int. J. Geo Inf..

[34]  J. Stewart,et al.  Development of VS Profile Database and Proxy‐Based Models for VS30 Prediction in the Pacific Northwest Region of North America , 2017 .

[35]  Shunichi Koshimura,et al.  The Potential Role of News Media to Construct a Machine Learning Based Damage Mapping Framework , 2021, Remote. Sens..

[36]  Chang-Guk Sun,et al.  Geologic site conditions and site coefficients for estimating earthquake ground motions in the inland areas of Korea , 2005 .

[37]  Raymond B. Seed,et al.  New Site Coefficients and Site Classification System Used in Recent Building Seismic Code Provisions , 2000 .

[38]  L. Faenza,et al.  ShakeMap implementation in Italy , 2008 .

[39]  Alexander Derendyaev,et al.  Machine Learning Methods for Seismic Hazards Forecast , 2019, Geosciences.

[40]  N. Theodulidis,et al.  Horizontal-to-vertical spectral ratio and geological conditions: The case of Garner Valley Downhole Array in southern California , 1996, Bulletin of the Seismological Society of America.

[41]  Lei Zhao,et al.  Continental crustal evolution and synchronous metallogeny through time in the North China Craton , 2020 .

[42]  Shie-Yui Liong,et al.  FLOOD STAGE FORECASTING WITH SUPPORT VECTOR MACHINES 1 , 2002 .

[43]  Philippe De Smedt,et al.  Towards a three-dimensional cost-effective registration of the archaeological heritage , 2013 .

[44]  Chang-Guk Sun,et al.  Geostatistical assessment for the regional zonation of seismic site effects in a coastal urban area using a GIS framework , 2016, Bulletin of Earthquake Engineering.

[45]  Chang-Guk Sun,et al.  Geospatial Big Data-Based Geostatistical Zonation of Seismic Site Effects in Seoul Metropolitan Area , 2017, ISPRS Int. J. Geo Inf..

[46]  David Carver,et al.  Nonlinear and Linear Site Response and Basin Effects in Seattle for the M 6.8 Nisqually, Washington, Earthquake , 2002 .

[47]  Wadii Boulila,et al.  A Novel Privacy Approach of Digital Aerial Images Based on Mersenne Twister Method with DNA Genetic Encoding and Chaos , 2020, Remote. Sens..

[48]  Katerina Kavoura,et al.  Assessment of Earthquake-Induced Landslide Hazard in Greece: From Arias Intensity to Spatial Distribution of Slope Resistance Demand , 2016 .

[49]  S. Myers,et al.  Body‐Wave Methods of Distinguishing between Explosions, Collapses, and Earthquakes: Application to Recent Events in North Korea , 2018, Seismological Research Letters.

[50]  D. Bui,et al.  Mapping wind erosion hazard with regression-based machine learning algorithms , 2020, Scientific Reports.

[51]  Site Response, Shallow Shear-Wave Velocity, and Damage in Los Gatos, California, from the 1989 Loma Prieta Earthquake , 2001 .

[52]  G. Heuvelink,et al.  A generic framework for spatial prediction of soil variables based on regression-kriging , 2004 .

[53]  Chang-Guk Sun,et al.  Spatial zonations for regional assessment of seismic site effects in the Seoul metropolitan area , 2014 .

[54]  R. Ulusay,et al.  Empirical correlations between shear wave velocity and penetration resistance for ground shaking assessments , 2007 .

[55]  D. Wald,et al.  Review Article Topographic Slope as a Proxy for Seismic Site Conditions and Amplification , 2007 .

[56]  Chang-Guk Sun,et al.  Correlations Between Shear Wave Velocity and In-Situ Penetration Test Results for Korean Soil Deposits , 2013, Pure and Applied Geophysics.

[57]  Aaron Klein,et al.  Auto-sklearn: Efficient and Robust Automated Machine Learning , 2019, Automated Machine Learning.