Using geoelectric field skewness and kurtosis to forecast the 2016/2/6, ML 6.6 Meinong, Taiwan Earthquake

  • Author(s): Hong-Jia Chen, Chien-Chih Chen, Guy Ouillon, and Didier Sornette
  • DOI: 10.3319/TAO.2016.11.01.01
  • Keywords: Earthquake precursors, geoelectric field, skewness, kurtosis, binary classification
  • Citation: Chen, H.-J., C.-C. Chen, G. Ouillon, and D. Sornette, 2017: Using geoelectric field skewness and kurtosis to forecast the 2016/2/6, ML 6.6 Meinong, Taiwan Earthquake. Terr. Atmos. Ocean. Sci., 28, 745-761, doi: 10.3319/TAO.2016.11.01.01
  • Treating geoelectric field distributions as earthquake precursors
  • Anomalous skewness and kurtosis of geoelectric fields before the Meinong earthquake
  • The GEMSTIP model can forecast the Meinong earthquake to some extent
Abstract

The earthquake-alarm model developed by Chen and Chen (Nat. Hazards 2016) is investigated to validate its forecasting performance for the 2016/2/6, ML 6.6 Meinong, Taiwan earthquake. This alarm model is based on geoelectric field skewness and kurtosis anomalies. The model parameters, such as the detection range and predicted time window, allow us to estimate the empirical relationships between geoelectric anomalies and large earthquakes. As a result, the skewness and kurtosis anomalies are shown to appear before the Meinong earthquake on the four neighboring stations (LIOQ, WANL, KAOH, and CHCH). According to the model analysis a time lag exists between anomaly clusters and earthquakes, depending on local geological features, as well as the durations over which anomalies are continuously observed, which might also display time dependence. In conclusion, this alarm model is able to correlate earthquakes and geoelectrical anomalies, with promising usefulness in forecasting large earthquakes.

Read 1018 times