An assessment on the use of bivariate, multivariate and soft computing techniques for collapse susceptibility in GIS environ

Creative Commons License

YILMAZ I., Marschalko M., Bednarik M.

JOURNAL OF EARTH SYSTEM SCIENCE, cilt.122, sa.2, ss.371-388, 2013 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 122 Sayı: 2
  • Basım Tarihi: 2013
  • Doi Numarası: 10.1007/s12040-013-0281-3
  • Dergi Adı: JOURNAL OF EARTH SYSTEM SCIENCE
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Sayfa Sayıları: ss.371-388
  • Anahtar Kelimeler: Collapse susceptibility map, gypsum, GIS, bivariate (conditional probability), multivariate (logistic regression), soft computing (artificial neural networks), PROBABILISTIC LANDSLIDE SUSCEPTIBILITY, ARTIFICIAL NEURAL-NETWORK, LOGISTIC-REGRESSION, GYPSUM, KARST, SIVAS, SINKHOLES, ZONATION, EXAMPLE, TURKEY
  • Sivas Cumhuriyet Üniversitesi Adresli: Evet

Özet

The paper presented herein compares and discusses the use of bivariate, multivariate and soft computing techniques for collapse susceptibility modelling. Conditional probability (CP), logistic regression (LR) and artificial neural networks (ANN) models representing the bivariate, multivariate and soft computing techniques were used in GIS based collapse susceptibility mapping in an area from Sivas basin (Turkey). Collapse-related factors, directly or indirectly related to the causes of collapse occurrence, such as distance from faults, slope angle and aspect, topographical elevation, distance from drainage, topographic wetness index (TWI), stream power index (SPI), Normalized Difference Vegetation Index (NDVI) by means of vegetation cover, distance from roads and settlements were used in the collapse susceptibility analyses. In the last stage of the analyses, collapse susceptibility maps were produced from the models, and they were then compared by means of their validations. However, Area Under Curve (AUC) values obtained from all three models showed that the map obtained from soft computing (ANN) model looks like more accurate than the other models, accuracies of all three models can be evaluated relatively similar. The results also showed that the conditional probability is an essential method in preparation of collapse susceptibility map and highly compatible with GIS operating features.