Comparison of indicator cokriging to disjunctive kriging for earthquake parametric estimation
Indicator cokriging and disjunctive kriging provide a regionalized variables approach for the simultaneous estimation of earthquake ground motion and the probability that certain ground motion levels will be exceeded. This has been the first application of indicator cokriging since this technique was created. Indicator cokriging (ICK) is demonstrated to be superior to disjunctive kriging (DK) for estimation of the exceedance probabilities, especially for those associated with higher level of ground motions. ICK provides a nonparametric estimation of primary variable with better local variability for clustered and scattered ground motion data. ICK also offers the flexibilities to handle both local and global estimation and yields accurate estimates when earthquake ground motion data are non-stationary and relatively stationary. DK estimation usually requires data stationarity or quasi-stationarity and the assumption that the data are univariate and bivariate normally distributed. DK provides accurate estimates when these assumptions are valid for a type of data and yields superior estimates for local estimation of quasi-stationary instrumented data and for modified Mercalli intensity data. The technique can be implemented with less computational effort and equivalent accuracy relative to indicator cokriging.
- Research Organization:
- Missouri Univ., Rolla (USA)
- OSTI ID:
- 6409250
- Country of Publication:
- United States
- Language:
- English
Similar Records
Comparison of disjunctive kriging to generalized probability kriging in application to the estimation of simulated and real data
Comparison of two techniques for applying disjunctive kriging: the Gaussian anamorphosis model versus the direct statistical inference of the bivariate distributions